Add files using upload-large-folder tool

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_ctc_loss_backward.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 <optional>
+
+
+
+#include <ATen/ops/_ctc_loss_backward_ops.h>
+
+namespace at {
+
+
+// aten::_ctc_loss_backward(Tensor grad, Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, Tensor neg_log_likelihood, Tensor log_alpha, int blank, bool zero_infinity=False) -> Tensor
+inline 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) {
+    return at::_ops::_ctc_loss_backward::call(grad, log_probs, targets, input_lengths, target_lengths, neg_log_likelihood, log_alpha, blank, zero_infinity);
+}
+
+// aten::_ctc_loss_backward.Tensor(Tensor grad, Tensor log_probs, Tensor targets, Tensor input_lengths, Tensor target_lengths, Tensor neg_log_likelihood, Tensor log_alpha, int blank, bool zero_infinity=False) -> Tensor
+inline 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) {
+    return at::_ops::_ctc_loss_backward_Tensor::call(grad, log_probs, targets, input_lengths, target_lengths, neg_log_likelihood, log_alpha, blank, zero_infinity);
+}
+
+// aten::_ctc_loss_backward.out(Tensor grad, Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, Tensor neg_log_likelihood, Tensor log_alpha, int blank, bool zero_infinity=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _ctc_loss_backward_out(at::Tensor & out, 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) {
+    return at::_ops::_ctc_loss_backward_out::call(grad, log_probs, targets, input_lengths, target_lengths, neg_log_likelihood, log_alpha, blank, zero_infinity, out);
+}
+// aten::_ctc_loss_backward.out(Tensor grad, Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, Tensor neg_log_likelihood, Tensor log_alpha, int blank, bool zero_infinity=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _ctc_loss_backward_outf(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, at::Tensor & out) {
+    return at::_ops::_ctc_loss_backward_out::call(grad, log_probs, targets, input_lengths, target_lengths, neg_log_likelihood, log_alpha, blank, zero_infinity, out);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_fake_quantize_learnable_per_channel_affine_backward.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 <optional>
+
+
+
+#include <ATen/ops/_fake_quantize_learnable_per_channel_affine_backward_ops.h>
+
+namespace at {
+
+
+// aten::_fake_quantize_learnable_per_channel_affine_backward(Tensor grad, Tensor self, Tensor scale, Tensor zero_point, int axis, int quant_min, int quant_max, float grad_factor=1.0) -> (Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _fake_quantize_learnable_per_channel_affine_backward(const at::Tensor & grad, const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis, int64_t quant_min, int64_t quant_max, double grad_factor=1.0) {
+    return at::_ops::_fake_quantize_learnable_per_channel_affine_backward::call(grad, self, scale, zero_point, axis, quant_min, quant_max, grad_factor);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_addcmul_ops.h

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_asin_ops.h

@@ -0,0 +1,50 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _foreach_asin {
+  using schema = ::std::vector<at::Tensor> (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_asin")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_asin(Tensor[] self) -> Tensor[]")
+  static ::std::vector<at::Tensor> call(at::TensorList self);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_asin_ {
+  using schema = void (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_asin_")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_asin_(Tensor(a!)[] self) -> ()")
+  static void call(at::TensorList self);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_asin_out {
+  using schema = void (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_asin")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_asin.out(Tensor[] self, *, Tensor(a!)[] out) -> ()")
+  static void call(at::TensorList self, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList out);
+};
+
+}} // namespace at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_sin_ops.h

@@ -0,0 +1,50 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _foreach_sin {
+  using schema = ::std::vector<at::Tensor> (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_sin")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_sin(Tensor[] self) -> Tensor[]")
+  static ::std::vector<at::Tensor> call(at::TensorList self);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_sin_ {
+  using schema = void (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_sin_")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_sin_(Tensor(a!)[] self) -> ()")
+  static void call(at::TensorList self);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_sin_out {
+  using schema = void (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_sin")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_sin.out(Tensor[] self, *, Tensor(a!)[] out) -> ()")
+  static void call(at::TensorList self, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList out);
+};
+
+}} // namespace at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_fw_primal_compositeexplicitautograd_dispatch.h

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_new_zeros_with_same_feature_meta_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 _new_zeros_with_same_feature_meta {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_new_zeros_with_same_feature_meta")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_new_zeros_with_same_feature_meta(Tensor self, Tensor other, *, int self_num_batch_dims=0) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & other, int64_t self_num_batch_dims);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, int64_t self_num_batch_dims);
+};
+
+struct TORCH_API _new_zeros_with_same_feature_meta_out {
+  using schema = at::Tensor & (const 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::_new_zeros_with_same_feature_meta")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_new_zeros_with_same_feature_meta.out(Tensor self, Tensor other, *, int self_num_batch_dims=0, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & other, int64_t self_num_batch_dims, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, int64_t self_num_batch_dims, at::Tensor & out);
+};
+
+}} // namespace at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_pdist_backward_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 & _pdist_backward_out(at::Tensor & out, const at::Tensor & grad, const at::Tensor & self, double p, const at::Tensor & pdist);
+TORCH_API at::Tensor & _pdist_backward_outf(const at::Tensor & grad, const at::Tensor & self, double p, const at::Tensor & pdist, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_scaled_dot_product_flash_attention_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 <optional>
+#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,c10::SymInt,c10::SymInt,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_flash_attention_cuda(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, double dropout_p=0.0, bool is_causal=false, bool return_debug_mask=false, ::std::optional<double> scale=::std::nullopt);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,c10::SymInt,c10::SymInt,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_flash_attention_nestedtensor_cuda(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, double dropout_p=0.0, bool is_causal=false, bool return_debug_mask=false, ::std::optional<double> scale=::std::nullopt);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_sparse_sum_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 _sparse_sum(const at::Tensor & self);
+TORCH_API at::Tensor _sparse_sum(const at::Tensor & self, at::ScalarType dtype);
+TORCH_API at::Tensor _sparse_sum(const at::Tensor & self, at::IntArrayRef dim, at::ScalarType dtype);
+
+} // namespace compositeimplicitautograd
+} // namespace at

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_bilinear2d_aa_meta_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 meta {
+
+TORCH_API at::Tensor _upsample_bilinear2d_aa(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor _upsample_bilinear2d_aa_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & _upsample_bilinear2d_aa_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & _upsample_bilinear2d_aa_outf(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+TORCH_API at::Tensor & _upsample_bilinear2d_aa_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & _upsample_bilinear2d_aa_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+
+} // namespace meta
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_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(const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales=::std::nullopt);
+TORCH_API at::Tensor _upsample_nearest_exact1d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales=::std::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact1d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales=::std::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact1d_outf(const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales, at::Tensor & out);
+TORCH_API at::Tensor & _upsample_nearest_exact1d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales=::std::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact1d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_wrapped_linear_prepack_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 _wrapped_linear_prepack(const at::Tensor & weight, const at::Tensor & weight_scale, const at::Tensor & weight_zero_point, const at::Tensor & bias);
+
+} // namespace compositeimplicitautograd
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/alias_copy_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 <optional>
+#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 & alias_copy_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor alias_copy(const at::Tensor & self);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/aminmax_cuda_dispatch.h

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/arccos_compositeimplicitautograd_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 compositeimplicitautograd {
+
+TORCH_API at::Tensor arccos(const at::Tensor & self);
+TORCH_API at::Tensor & arccos_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & arccos_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & arccos_(at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/arcsinh.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 <optional>
+
+
+
+#include <ATen/ops/arcsinh_ops.h>
+
+namespace at {
+
+
+// aten::arcsinh(Tensor self) -> Tensor
+inline at::Tensor arcsinh(const at::Tensor & self) {
+    return at::_ops::arcsinh::call(self);
+}
+
+// aten::arcsinh_(Tensor(a!) self) -> Tensor(a!)
+inline at::Tensor & arcsinh_(at::Tensor & self) {
+    return at::_ops::arcsinh_::call(self);
+}
+
+// aten::arcsinh.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & arcsinh_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::arcsinh_out::call(self, out);
+}
+// aten::arcsinh.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & arcsinh_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::arcsinh_out::call(self, out);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/atan2_ops.h

@@ -0,0 +1,50 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API atan2_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::atan2")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "atan2.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+};
+
+struct TORCH_API atan2_ {
+  using schema = at::Tensor & (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::atan2_")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "atan2_(Tensor(a!) self, Tensor other) -> Tensor(a!)")
+  static at::Tensor & call(at::Tensor & self, const at::Tensor & other);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Tensor & other);
+};
+
+struct TORCH_API atan2 {
+  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::atan2")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "atan2(Tensor self, Tensor other) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & other);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other);
+};
+
+}} // namespace at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/avg_pool3d_backward_cuda_dispatch.h

@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor avg_pool3d_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, ::std::optional<int64_t> divisor_override);
+TORCH_API at::Tensor & avg_pool3d_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, ::std::optional<int64_t> divisor_override);
+TORCH_API at::Tensor & avg_pool3d_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, ::std::optional<int64_t> divisor_override, at::Tensor & grad_input);
+
+} // namespace cuda
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/bernoulli_native.h

@@ -0,0 +1,28 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#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 bernoulli(const at::Tensor & self, ::std::optional<at::Generator> generator=::std::nullopt);
+TORCH_API at::Tensor & bernoulli_out(const at::Tensor & self, ::std::optional<at::Generator> generator, at::Tensor & out);
+TORCH_API at::Tensor bernoulli(const at::Tensor & self, const at::Tensor & p, ::std::optional<at::Generator> generator=::std::nullopt);
+TORCH_API at::Tensor & bernoulli_Tensor_out(const at::Tensor & self, const at::Tensor & p, ::std::optional<at::Generator> generator, at::Tensor & out);
+TORCH_API at::Tensor & bernoulli_(at::Tensor & self, const at::Tensor & p, ::std::optional<at::Generator> generator=::std::nullopt);
+TORCH_API at::Tensor & bernoulli_float_out(const at::Tensor & self, double p, ::std::optional<at::Generator> generator, at::Tensor & out);
+TORCH_API at::Tensor & bernoulli_(at::Tensor & self, double p=0.5, ::std::optional<at::Generator> generator=::std::nullopt);
+TORCH_API at::Tensor bernoulli(const at::Tensor & self, double p, ::std::optional<at::Generator> generator=::std::nullopt);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/broadcast_to_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 <optional>
+#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 broadcast_to_symint(const at::Tensor & self, c10::SymIntArrayRef size);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/bucketize_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 & bucketize_out(at::Tensor & out, const at::Scalar & self, const at::Tensor & boundaries, bool out_int32=false, bool right=false);
+TORCH_API at::Tensor & bucketize_outf(const at::Scalar & self, const at::Tensor & boundaries, bool out_int32, bool right, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/clamp_max_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_max(const at::Tensor & self, const at::Scalar & max);
+TORCH_API at::Tensor & clamp_max_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & max);
+TORCH_API at::Tensor & clamp_max_outf(const at::Tensor & self, const at::Scalar & max, at::Tensor & out);
+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_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & max);
+TORCH_API at::Tensor & clamp_max_outf(const at::Tensor & self, const at::Tensor & max, at::Tensor & out);
+TORCH_API at::Tensor & clamp_max_(at::Tensor & self, const at::Tensor & max);
+
+} // namespace meta
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/clamp_min.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 <optional>
+
+
+
+#include <ATen/ops/clamp_min_ops.h>
+
+namespace at {
+
+
+// aten::clamp_min(Tensor self, Scalar min) -> Tensor
+inline at::Tensor clamp_min(const at::Tensor & self, const at::Scalar & min) {
+    return at::_ops::clamp_min::call(self, min);
+}
+
+// aten::clamp_min.Tensor(Tensor self, Tensor min) -> Tensor
+inline at::Tensor clamp_min(const at::Tensor & self, const at::Tensor & min) {
+    return at::_ops::clamp_min_Tensor::call(self, min);
+}
+
+// aten::clamp_min_(Tensor(a!) self, Scalar min) -> Tensor(a!)
+inline at::Tensor & clamp_min_(at::Tensor & self, const at::Scalar & min) {
+    return at::_ops::clamp_min_::call(self, min);
+}
+
+// aten::clamp_min_.Tensor(Tensor(a!) self, Tensor min) -> Tensor(a!)
+inline at::Tensor & clamp_min_(at::Tensor & self, const at::Tensor & min) {
+    return at::_ops::clamp_min__Tensor::call(self, min);
+}
+
+// aten::clamp_min.out(Tensor self, Scalar min, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & clamp_min_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & min) {
+    return at::_ops::clamp_min_out::call(self, min, out);
+}
+// aten::clamp_min.out(Tensor self, Scalar min, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & clamp_min_outf(const at::Tensor & self, const at::Scalar & min, at::Tensor & out) {
+    return at::_ops::clamp_min_out::call(self, min, out);
+}
+
+// aten::clamp_min.Tensor_out(Tensor self, Tensor min, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & clamp_min_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & min) {
+    return at::_ops::clamp_min_Tensor_out::call(self, min, out);
+}
+// aten::clamp_min.Tensor_out(Tensor self, Tensor min, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & clamp_min_outf(const at::Tensor & self, const at::Tensor & min, at::Tensor & out) {
+    return at::_ops::clamp_min_Tensor_out::call(self, min, out);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/constant_pad_nd.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 <optional>
+
+
+
+#include <ATen/ops/constant_pad_nd_ops.h>
+
+namespace at {
+
+
+// aten::constant_pad_nd(Tensor self, SymInt[] pad, Scalar value=0) -> Tensor
+inline at::Tensor constant_pad_nd(const at::Tensor & self, at::IntArrayRef pad, const at::Scalar & value=0) {
+    return at::_ops::constant_pad_nd::call(self, c10::fromIntArrayRefSlow(pad), value);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor constant_pad_nd(const at::Tensor & self, at::IntArrayRef pad, const at::Scalar & value=0) {
+    return at::_ops::constant_pad_nd::call(self, c10::fromIntArrayRefSlow(pad), value);
+  }
+}
+
+// aten::constant_pad_nd(Tensor self, SymInt[] pad, Scalar value=0) -> Tensor
+inline at::Tensor constant_pad_nd_symint(const at::Tensor & self, c10::SymIntArrayRef pad, const at::Scalar & value=0) {
+    return at::_ops::constant_pad_nd::call(self, pad, value);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor constant_pad_nd(const at::Tensor & self, c10::SymIntArrayRef pad, const at::Scalar & value=0) {
+    return at::_ops::constant_pad_nd::call(self, pad, value);
+  }
+}
+
+// aten::constant_pad_nd.out(Tensor self, SymInt[] pad, Scalar value=0, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & constant_pad_nd_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef pad, const at::Scalar & value=0) {
+    return at::_ops::constant_pad_nd_out::call(self, c10::fromIntArrayRefSlow(pad), value, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & constant_pad_nd_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef pad, const at::Scalar & value=0) {
+    return at::_ops::constant_pad_nd_out::call(self, c10::fromIntArrayRefSlow(pad), value, out);
+  }
+}
+
+// aten::constant_pad_nd.out(Tensor self, SymInt[] pad, Scalar value=0, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & constant_pad_nd_outf(const at::Tensor & self, at::IntArrayRef pad, const at::Scalar & value, at::Tensor & out) {
+    return at::_ops::constant_pad_nd_out::call(self, c10::fromIntArrayRefSlow(pad), value, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & constant_pad_nd_outf(const at::Tensor & self, at::IntArrayRef pad, const at::Scalar & value, at::Tensor & out) {
+    return at::_ops::constant_pad_nd_out::call(self, c10::fromIntArrayRefSlow(pad), value, out);
+  }
+}
+
+// aten::constant_pad_nd.out(Tensor self, SymInt[] pad, Scalar value=0, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & constant_pad_nd_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef pad, const at::Scalar & value=0) {
+    return at::_ops::constant_pad_nd_out::call(self, pad, value, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & constant_pad_nd_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef pad, const at::Scalar & value=0) {
+    return at::_ops::constant_pad_nd_out::call(self, pad, value, out);
+  }
+}
+
+// aten::constant_pad_nd.out(Tensor self, SymInt[] pad, Scalar value=0, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & constant_pad_nd_symint_outf(const at::Tensor & self, c10::SymIntArrayRef pad, const at::Scalar & value, at::Tensor & out) {
+    return at::_ops::constant_pad_nd_out::call(self, pad, value, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & constant_pad_nd_outf(const at::Tensor & self, c10::SymIntArrayRef pad, const at::Scalar & value, at::Tensor & out) {
+    return at::_ops::constant_pad_nd_out::call(self, pad, value, out);
+  }
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/conv1d.h

@@ -0,0 +1,69 @@
+#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 <optional>
+
+
+
+#include <ATen/ops/conv1d_ops.h>
+
+namespace at {
+
+
+// aten::conv1d(Tensor input, Tensor weight, Tensor? bias=None, SymInt[1] stride=1, SymInt[1] padding=0, SymInt[1] dilation=1, SymInt groups=1) -> Tensor
+inline at::Tensor conv1d(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias={}, at::IntArrayRef stride=1, at::IntArrayRef padding=0, at::IntArrayRef dilation=1, int64_t groups=1) {
+    return at::_ops::conv1d::call(input, weight, bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation), groups);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor conv1d(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias={}, at::IntArrayRef stride=1, at::IntArrayRef padding=0, at::IntArrayRef dilation=1, int64_t groups=1) {
+    return at::_ops::conv1d::call(input, weight, bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation), groups);
+  }
+}
+
+// aten::conv1d(Tensor input, Tensor weight, Tensor? bias=None, SymInt[1] stride=1, SymInt[1] padding=0, SymInt[1] dilation=1, SymInt groups=1) -> Tensor
+inline at::Tensor conv1d_symint(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias={}, c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef dilation=c10::SymInt(1), c10::SymInt groups=1) {
+    return at::_ops::conv1d::call(input, weight, bias, stride, padding, dilation, groups);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor conv1d(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias={}, c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef dilation=c10::SymInt(1), c10::SymInt groups=1) {
+    return at::_ops::conv1d::call(input, weight, bias, stride, padding, dilation, groups);
+  }
+}
+
+// aten::conv1d.padding(Tensor input, Tensor weight, Tensor? bias=None, SymInt[1] stride=1, str padding="valid", SymInt[1] dilation=1, SymInt groups=1) -> Tensor
+inline at::Tensor conv1d(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, c10::string_view padding, at::IntArrayRef dilation=1, int64_t groups=1) {
+    return at::_ops::conv1d_padding::call(input, weight, bias, c10::fromIntArrayRefSlow(stride), padding, c10::fromIntArrayRefSlow(dilation), groups);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor conv1d(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, c10::string_view padding, at::IntArrayRef dilation=1, int64_t groups=1) {
+    return at::_ops::conv1d_padding::call(input, weight, bias, c10::fromIntArrayRefSlow(stride), padding, c10::fromIntArrayRefSlow(dilation), groups);
+  }
+}
+
+// aten::conv1d.padding(Tensor input, Tensor weight, Tensor? bias=None, SymInt[1] stride=1, str padding="valid", SymInt[1] dilation=1, SymInt groups=1) -> Tensor
+inline at::Tensor conv1d_symint(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::string_view padding, c10::SymIntArrayRef dilation=c10::SymInt(1), c10::SymInt groups=1) {
+    return at::_ops::conv1d_padding::call(input, weight, bias, stride, padding, dilation, groups);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor conv1d(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::string_view padding, c10::SymIntArrayRef dilation=c10::SymInt(1), c10::SymInt groups=1) {
+    return at::_ops::conv1d_padding::call(input, weight, bias, stride, padding, dilation, groups);
+  }
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/conv_depthwise3d.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 <optional>
+
+
+
+#include <ATen/ops/conv_depthwise3d_ops.h>
+
+namespace at {
+
+
+// aten::conv_depthwise3d(Tensor self, Tensor weight, SymInt[3] kernel_size, Tensor? bias, SymInt[3] stride, SymInt[3] padding, SymInt[3] dilation) -> Tensor
+inline at::Tensor conv_depthwise3d(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation) {
+    return at::_ops::conv_depthwise3d::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation));
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor conv_depthwise3d(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation) {
+    return at::_ops::conv_depthwise3d::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation));
+  }
+}
+
+// aten::conv_depthwise3d(Tensor self, Tensor weight, SymInt[3] kernel_size, Tensor? bias, SymInt[3] stride, SymInt[3] padding, SymInt[3] dilation) -> Tensor
+inline at::Tensor conv_depthwise3d_symint(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation) {
+    return at::_ops::conv_depthwise3d::call(self, weight, kernel_size, bias, stride, padding, dilation);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor conv_depthwise3d(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation) {
+    return at::_ops::conv_depthwise3d::call(self, weight, kernel_size, bias, stride, padding, dilation);
+  }
+}
+
+// aten::conv_depthwise3d.out(Tensor self, Tensor weight, SymInt[3] kernel_size, Tensor? bias, SymInt[3] stride, SymInt[3] padding, SymInt[3] dilation, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & conv_depthwise3d_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation) {
+    return at::_ops::conv_depthwise3d_out::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & conv_depthwise3d_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation) {
+    return at::_ops::conv_depthwise3d_out::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation), out);
+  }
+}
+
+// aten::conv_depthwise3d.out(Tensor self, Tensor weight, SymInt[3] kernel_size, Tensor? bias, SymInt[3] stride, SymInt[3] padding, SymInt[3] dilation, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & conv_depthwise3d_outf(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, at::Tensor & out) {
+    return at::_ops::conv_depthwise3d_out::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & conv_depthwise3d_outf(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, at::Tensor & out) {
+    return at::_ops::conv_depthwise3d_out::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation), out);
+  }
+}
+
+// aten::conv_depthwise3d.out(Tensor self, Tensor weight, SymInt[3] kernel_size, Tensor? bias, SymInt[3] stride, SymInt[3] padding, SymInt[3] dilation, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & conv_depthwise3d_symint_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation) {
+    return at::_ops::conv_depthwise3d_out::call(self, weight, kernel_size, bias, stride, padding, dilation, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & conv_depthwise3d_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation) {
+    return at::_ops::conv_depthwise3d_out::call(self, weight, kernel_size, bias, stride, padding, dilation, out);
+  }
+}
+
+// aten::conv_depthwise3d.out(Tensor self, Tensor weight, SymInt[3] kernel_size, Tensor? bias, SymInt[3] stride, SymInt[3] padding, SymInt[3] dilation, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & conv_depthwise3d_symint_outf(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, at::Tensor & out) {
+    return at::_ops::conv_depthwise3d_out::call(self, weight, kernel_size, bias, stride, padding, dilation, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & conv_depthwise3d_outf(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, at::Tensor & out) {
+    return at::_ops::conv_depthwise3d_out::call(self, weight, kernel_size, bias, stride, padding, dilation, out);
+  }
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/diagflat_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 <optional>
+#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 diagflat(const at::Tensor & self, int64_t offset=0);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/dropout.h

@@ -0,0 +1,35 @@
+#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 <optional>
+
+
+
+#include <ATen/ops/dropout_ops.h>
+
+namespace at {
+
+
+// aten::dropout(Tensor input, float p, bool train) -> Tensor
+inline at::Tensor dropout(const at::Tensor & input, double p, bool train) {
+    return at::_ops::dropout::call(input, p, train);
+}
+
+// aten::dropout_(Tensor(a!) self, float p, bool train) -> Tensor(a!)
+inline at::Tensor & dropout_(at::Tensor & self, double p, bool train) {
+    return at::_ops::dropout_::call(self, p, train);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/erf_ops.h

@@ -0,0 +1,50 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API erf {
+  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::erf")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "erf(Tensor self) -> Tensor")
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+struct TORCH_API erf_ {
+  using schema = 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::erf_")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "erf_(Tensor(a!) self) -> Tensor(a!)")
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API erf_out {
+  using schema = 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::erf")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "erf.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Tensor & out);
+};
+
+}} // namespace at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/expand_as_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 <optional>
+#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 expand_as(const at::Tensor & self, const at::Tensor & other);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fft_hfft_compositeimplicitautograd_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 compositeimplicitautograd {
+
+TORCH_API at::Tensor fft_hfft(const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor fft_hfft_symint(const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_hfft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_hfft_outf(const at::Tensor & self, ::std::optional<int64_t> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+TORCH_API at::Tensor & fft_hfft_symint_out(at::Tensor & out, const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_hfft_symint_outf(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fft_ihfft.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 <optional>
+
+
+
+#include <ATen/ops/fft_ihfft_ops.h>
+
+namespace at {
+
+
+// aten::fft_ihfft(Tensor self, SymInt? n=None, int dim=-1, str? norm=None) -> Tensor
+inline at::Tensor fft_ihfft(const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor fft_ihfft(const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm);
+  }
+}
+
+// aten::fft_ihfft(Tensor self, SymInt? n=None, int dim=-1, str? norm=None) -> Tensor
+inline at::Tensor fft_ihfft_symint(const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft::call(self, n, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor fft_ihfft(const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft::call(self, n, dim, norm);
+  }
+}
+
+// aten::fft_ihfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & fft_ihfft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_ihfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft_outf(const at::Tensor & self, ::std::optional<int64_t> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & fft_ihfft_outf(const at::Tensor & self, ::std::optional<int64_t> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_ihfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft_symint_out(at::Tensor & out, const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft_out::call(self, n, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & fft_ihfft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft_out::call(self, n, dim, norm, out);
+  }
+}
+
+// aten::fft_ihfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft_symint_outf(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft_out::call(self, n, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & fft_ihfft_outf(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft_out::call(self, n, dim, norm, out);
+  }
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/full.h

@@ -0,0 +1,131 @@
+#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 <optional>
+
+
+
+#include <ATen/ops/full_ops.h>
+
+namespace at {
+
+
+// aten::full.names(int[] size, Scalar fill_value, *, Dimname[]? names, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor full(at::IntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::DimnameList> names, at::TensorOptions options={}) {
+    return at::_ops::full_names::call(size, fill_value, names, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+// aten::full.names(int[] size, Scalar fill_value, *, Dimname[]? names, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor full(at::IntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::DimnameList> names, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::full_names::call(size, fill_value, names, dtype, layout, device, pin_memory);
+}
+
+// aten::full(SymInt[] size, Scalar fill_value, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor full(at::IntArrayRef size, const at::Scalar & fill_value, at::TensorOptions options={}) {
+    return at::_ops::full::call(c10::fromIntArrayRefSlow(size), fill_value, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor full(at::IntArrayRef size, const at::Scalar & fill_value, at::TensorOptions options={}) {
+    return at::_ops::full::call(c10::fromIntArrayRefSlow(size), fill_value, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+  }
+}
+
+// aten::full(SymInt[] size, Scalar fill_value, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor full(at::IntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::full::call(c10::fromIntArrayRefSlow(size), fill_value, dtype, layout, device, pin_memory);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor full(at::IntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::full::call(c10::fromIntArrayRefSlow(size), fill_value, dtype, layout, device, pin_memory);
+  }
+}
+
+// aten::full(SymInt[] size, Scalar fill_value, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor full_symint(c10::SymIntArrayRef size, const at::Scalar & fill_value, at::TensorOptions options={}) {
+    return at::_ops::full::call(size, fill_value, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor full(c10::SymIntArrayRef size, const at::Scalar & fill_value, at::TensorOptions options={}) {
+    return at::_ops::full::call(size, fill_value, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+  }
+}
+
+// aten::full(SymInt[] size, Scalar fill_value, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor full_symint(c10::SymIntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::full::call(size, fill_value, dtype, layout, device, pin_memory);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor full(c10::SymIntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::full::call(size, fill_value, dtype, layout, device, pin_memory);
+  }
+}
+
+// aten::full.out(SymInt[] size, Scalar fill_value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & full_out(at::Tensor & out, at::IntArrayRef size, const at::Scalar & fill_value) {
+    return at::_ops::full_out::call(c10::fromIntArrayRefSlow(size), fill_value, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & full_out(at::Tensor & out, at::IntArrayRef size, const at::Scalar & fill_value) {
+    return at::_ops::full_out::call(c10::fromIntArrayRefSlow(size), fill_value, out);
+  }
+}
+
+// aten::full.out(SymInt[] size, Scalar fill_value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & full_outf(at::IntArrayRef size, const at::Scalar & fill_value, at::Tensor & out) {
+    return at::_ops::full_out::call(c10::fromIntArrayRefSlow(size), fill_value, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & full_outf(at::IntArrayRef size, const at::Scalar & fill_value, at::Tensor & out) {
+    return at::_ops::full_out::call(c10::fromIntArrayRefSlow(size), fill_value, out);
+  }
+}
+
+// aten::full.out(SymInt[] size, Scalar fill_value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & full_symint_out(at::Tensor & out, c10::SymIntArrayRef size, const at::Scalar & fill_value) {
+    return at::_ops::full_out::call(size, fill_value, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & full_out(at::Tensor & out, c10::SymIntArrayRef size, const at::Scalar & fill_value) {
+    return at::_ops::full_out::call(size, fill_value, out);
+  }
+}
+
+// aten::full.out(SymInt[] size, Scalar fill_value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & full_symint_outf(c10::SymIntArrayRef size, const at::Scalar & fill_value, at::Tensor & out) {
+    return at::_ops::full_out::call(size, fill_value, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & full_outf(c10::SymIntArrayRef size, const at::Scalar & fill_value, at::Tensor & out) {
+    return at::_ops::full_out::call(size, fill_value, out);
+  }
+}
+
+// aten::full.names_out(int[] size, Scalar fill_value, *, Dimname[]? names, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & full_out(at::Tensor & out, at::IntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::DimnameList> names) {
+    return at::_ops::full_names_out::call(size, fill_value, names, out);
+}
+// aten::full.names_out(int[] size, Scalar fill_value, *, Dimname[]? names, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & full_outf(at::IntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::DimnameList> names, at::Tensor & out) {
+    return at::_ops::full_names_out::call(size, fill_value, names, out);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/glu_jvp_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 & glu_jvp_out(at::Tensor & out, const at::Tensor & glu, const at::Tensor & x, const at::Tensor & dx, int64_t dim);
+TORCH_API at::Tensor & glu_jvp_outf(const at::Tensor & glu, const at::Tensor & x, const at::Tensor & dx, int64_t dim, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/gru_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 gru_input {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, at::TensorList, bool, int64_t, double, bool, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::gru")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "input")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "gru.input(Tensor input, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, bool batch_first) -> (Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & input, const at::Tensor & hx, at::TensorList params, bool has_biases, int64_t num_layers, double dropout, bool train, bool bidirectional, bool batch_first);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & hx, at::TensorList params, bool has_biases, int64_t num_layers, double dropout, bool train, bool bidirectional, bool batch_first);
+};
+
+struct TORCH_API gru_data {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, at::TensorList, bool, int64_t, double, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::gru")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "data")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "gru.data(Tensor data, Tensor batch_sizes, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional) -> (Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & data, const at::Tensor & batch_sizes, const at::Tensor & hx, at::TensorList params, bool has_biases, int64_t num_layers, double dropout, bool train, bool bidirectional);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & data, const at::Tensor & batch_sizes, const at::Tensor & hx, at::TensorList params, bool has_biases, int64_t num_layers, double dropout, bool train, bool bidirectional);
+};
+
+}} // namespace at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/hardshrink_backward_meta.h

@@ -0,0 +1,27 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeMetaFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#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_hardshrink_backward : public TensorIteratorBase {
+    
+    
+    void meta(const at::Tensor & grad_out, const at::Tensor & self, const at::Scalar & lambd);
+};
+
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/im2col_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 <optional>
+#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 im2col_cpu(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor & im2col_out_cpu(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride, at::Tensor & out);
+TORCH_API at::Tensor im2col_cuda(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor & im2col_out_cuda(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride, at::Tensor & out);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/is_complex_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 <optional>
+#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_complex(const at::Tensor & self);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/is_distributed_ops.h

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/isnan_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 isnan(const at::Tensor & self);
+
+} // namespace cpu
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/isreal_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 isreal {
+  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::isreal")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "isreal(Tensor self) -> Tensor")
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_lu_factor_ex_cpu_dispatch.h

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_matrix_exp_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 <optional>
+#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 & linalg_matrix_exp_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor linalg_matrix_exp(const at::Tensor & self);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_slogdet_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_slogdet {
+  using schema = ::std::tuple<at::Tensor,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::linalg_slogdet")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "linalg_slogdet(Tensor A) -> (Tensor sign, Tensor logabsdet)")
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & A);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & A);
+};
+
+struct TORCH_API linalg_slogdet_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (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::linalg_slogdet")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "linalg_slogdet.out(Tensor A, *, Tensor(a!) sign, Tensor(b!) logabsdet) -> (Tensor(a!) sign, Tensor(b!) logabsdet)")
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & A, at::Tensor & sign, at::Tensor & logabsdet);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & A, at::Tensor & sign, at::Tensor & logabsdet);
+};
+
+}} // namespace at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/max_pool2d_with_indices_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 <optional>
+#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_max_pool2d_with_indices : public at::impl::MetaBase {
+    
+    
+    void meta(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool ceil_mode);
+};
+
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/multilabel_margin_loss_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 multilabel_margin_loss_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & target, int64_t reduction, const at::Tensor & is_target);
+TORCH_API at::Tensor & multilabel_margin_loss_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & target, int64_t reduction, const at::Tensor & is_target);
+TORCH_API at::Tensor & multilabel_margin_loss_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & target, int64_t reduction, const at::Tensor & is_target, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at