diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_cast_Int_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_cast_Int_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..09a41c603cd966d6ddac40b5688cc63fad87899d
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_cast_Int_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_Int {
+  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_Int")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_cast_Int(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
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_coalesce_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_coalesce_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..26c3d1b66de0495df5eb2b63dff8901a0a9b8e29
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_coalesce_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 _coalesce {
+  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::_coalesce")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_coalesce(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 _coalesce_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::_coalesce")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_coalesce.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
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_conj_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_conj_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..d5cabe749e9b3fba9ace0fdd1e67408049559d60
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_conj_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 _conj {
+  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::_conj")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_conj(Tensor(a) self) -> Tensor(a)")
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_dirichlet_grad_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_dirichlet_grad_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..269157ec236d3ebdb7161bb33ddc72444b93dca9
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_dirichlet_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 _dirichlet_grad {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_dirichlet_grad")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_dirichlet_grad(Tensor x, Tensor alpha, Tensor total) -> Tensor")
+  static at::Tensor call(const at::Tensor & x, const at::Tensor & alpha, const at::Tensor & total);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & x, const at::Tensor & alpha, const at::Tensor & total);
+};
+
+struct TORCH_API _dirichlet_grad_out {
+  using schema = at::Tensor & (const 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::_dirichlet_grad")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_dirichlet_grad.out(Tensor x, Tensor alpha, Tensor total, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & x, const at::Tensor & alpha, const at::Tensor & total, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & x, const at::Tensor & alpha, const at::Tensor & total, at::Tensor & out);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_empty_per_channel_affine_quantized_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_empty_per_channel_affine_quantized_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6bc1b277f36cd3ffda42b4e5d583e64b2f4a1232
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_empty_per_channel_affine_quantized_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & _empty_per_channel_affine_quantized_out(at::Tensor & out, at::IntArrayRef size, const at::Tensor & scales, const at::Tensor & zero_points, int64_t axis, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous);
+TORCH_API at::Tensor & _empty_per_channel_affine_quantized_outf(at::IntArrayRef size, const at::Tensor & scales, const at::Tensor & zero_points, int64_t axis, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor & _empty_per_channel_affine_quantized_symint_out(at::Tensor & out, c10::SymIntArrayRef size, const at::Tensor & scales, const at::Tensor & zero_points, int64_t axis, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous);
+TORCH_API at::Tensor & _empty_per_channel_affine_quantized_symint_outf(c10::SymIntArrayRef size, const at::Tensor & scales, const at::Tensor & zero_points, int64_t axis, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_fill_mem_eff_dropout_mask_cuda_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_fill_mem_eff_dropout_mask_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..fb855f64fc0e254c13c73c4386adc540cdd89f54
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_fill_mem_eff_dropout_mask_cuda_dispatch.h
@@ -0,0 +1,23 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor & _fill_mem_eff_dropout_mask_(at::Tensor & self, double dropout_p, int64_t seed, int64_t offset);
+
+} // namespace cuda
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_acos.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_acos.h
new file mode 100644
index 0000000000000000000000000000000000000000..cb6f41460f8b3d436be714a43ccbed829dced2e1
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_acos.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/_foreach_acos_ops.h>
+
+namespace at {
+
+
+// aten::_foreach_acos(Tensor[] self) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_acos(at::TensorList self) {
+    return at::_ops::_foreach_acos::call(self);
+}
+
+// aten::_foreach_acos_(Tensor(a!)[] self) -> ()
+inline void _foreach_acos_(at::TensorList self) {
+    return at::_ops::_foreach_acos_::call(self);
+}
+
+// aten::_foreach_acos.out(Tensor[] self, *, Tensor(a!)[] out) -> ()
+inline void _foreach_acos_out(at::TensorList out, at::TensorList self) {
+    return at::_ops::_foreach_acos_out::call(self, out);
+}
+// aten::_foreach_acos.out(Tensor[] self, *, Tensor(a!)[] out) -> ()
+inline void _foreach_acos_outf(at::TensorList self, at::TensorList out) {
+    return at::_ops::_foreach_acos_out::call(self, out);
+}
+
+}
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_ceil_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_ceil_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..5e717be262fb563c527c9558d43f49407dc4a165
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_ceil_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_ceil {
+  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_ceil")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_ceil(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_ceil_ {
+  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_ceil_")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_ceil_(Tensor(a!)[] self) -> ()")
+  static void call(at::TensorList self);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_ceil_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_ceil")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_ceil.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
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_grid_sampler_2d_cpu_fallback_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_grid_sampler_2d_cpu_fallback_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..f5643f3400bda609b773d5819f7cdacbd7f554ff
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_grid_sampler_2d_cpu_fallback_ops.h
@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _grid_sampler_2d_cpu_fallback {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, int64_t, int64_t, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_grid_sampler_2d_cpu_fallback")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_grid_sampler_2d_cpu_fallback(Tensor input, Tensor grid, int interpolation_mode, int padding_mode, bool align_corners) -> Tensor")
+  static at::Tensor call(const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners);
+};
+
+struct TORCH_API _grid_sampler_2d_cpu_fallback_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, int64_t, int64_t, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_grid_sampler_2d_cpu_fallback")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_grid_sampler_2d_cpu_fallback.out(Tensor input, Tensor grid, int interpolation_mode, int padding_mode, bool align_corners, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners, at::Tensor & out);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_make_per_channel_quantized_tensor_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_make_per_channel_quantized_tensor_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..2a6f90dd87402c9c39151641e0e75aaaaa433f29
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_make_per_channel_quantized_tensor_native.h
@@ -0,0 +1,23 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <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 & _make_per_channel_quantized_tensor_out(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis, at::Tensor & out);
+TORCH_API at::Tensor make_per_channel_quantized_tensor_cpu(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis);
+TORCH_API at::Tensor make_per_channel_quantized_tensor_cuda(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_reshape_copy_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_reshape_copy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..f5343c0a03a950e23bbba71c21c1ffb9039fb123
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_reshape_copy_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 _reshape_copy_symint(const at::Tensor & self, c10::SymIntArrayRef size);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_scaled_dot_product_fused_attention_overrideable_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_scaled_dot_product_fused_attention_overrideable_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..d9b4eaad5d568a5e6217c19981c9f2a4890a8db9
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_scaled_dot_product_fused_attention_overrideable_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 _scaled_dot_product_fused_attention_overrideable {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,c10::SymInt,c10::SymInt,at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, double, bool, bool, ::std::optional<double>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_scaled_dot_product_fused_attention_overrideable")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_scaled_dot_product_fused_attention_overrideable(Tensor query, Tensor key, Tensor value, Tensor? attn_bias=None, float dropout_p=0.0, bool is_causal=False, bool return_debug_mask=False, *, float? scale=None) -> (Tensor output, Tensor logsumexp, Tensor cum_seq_q, Tensor cum_seq_k, SymInt max_q, SymInt max_k, Tensor philox_seed, Tensor philox_offset, Tensor debug_attn_mask)")
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,c10::SymInt,c10::SymInt,at::Tensor,at::Tensor,at::Tensor> call(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const ::std::optional<at::Tensor> & attn_bias, double dropout_p, bool is_causal, bool return_debug_mask, ::std::optional<double> scale);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,c10::SymInt,c10::SymInt,at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const ::std::optional<at::Tensor> & attn_bias, double dropout_p, bool is_causal, bool return_debug_mask, ::std::optional<double> scale);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_slow_conv2d_forward_cuda_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_slow_conv2d_forward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..06f62ad1aa23b79b7900c29f8ce03d1fae92102a
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_slow_conv2d_forward_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor _slow_conv2d_forward(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding);
+TORCH_API at::Tensor _slow_conv2d_forward_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);
+TORCH_API at::Tensor & _slow_conv2d_forward_out(at::Tensor & output, const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding);
+TORCH_API at::Tensor & _slow_conv2d_forward_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::Tensor & output);
+TORCH_API at::Tensor & _slow_conv2d_forward_symint_out(at::Tensor & output, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding);
+TORCH_API at::Tensor & _slow_conv2d_forward_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, at::Tensor & output);
+
+} // namespace cuda
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_test_optional_filled_intlist_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_test_optional_filled_intlist_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..640bc48116d66657d94fc859f5fa7c51c968ebf5
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_test_optional_filled_intlist_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 & _test_optional_filled_intlist_out(const at::Tensor & values, at::OptionalIntArrayRef addends, at::Tensor & out);
+TORCH_API at::Tensor _test_optional_intlist(const at::Tensor & values, at::OptionalIntArrayRef addends);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_test_serialization_subcmul_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_test_serialization_subcmul_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..eafd2b794b8163ea6f1199cbd74ad781edf7436d
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_test_serialization_subcmul_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 _test_serialization_subcmul(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+
+} // namespace compositeimplicitautograd
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_unsafe_index_put_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_unsafe_index_put_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..ecac6314ad0c210b5507c2577d9e7231ff7c0ba2
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_unsafe_index_put_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 _unsafe_index_put(const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate=false);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..433d069ede44ebd4964c7379c08cf5e4d27955b3
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,24 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautogradnonfunctional {
+
+TORCH_API at::Tensor _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);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_meta_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d3f82b0ded6cff40a9f01a84d39f2a9c8b7ba9db
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_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_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 meta
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_cuda_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e745633d2b8220d7ff64a970c75203d729fa7de3
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor _upsample_nearest_exact3d(const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales_d=::std::nullopt, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor _upsample_nearest_exact3d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales_d=::std::nullopt, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact3d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales_d=::std::nullopt, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact3d_outf(const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales_d, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+TORCH_API at::Tensor & _upsample_nearest_exact3d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales_d=::std::nullopt, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact3d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales_d, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/acos_meta.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/acos_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..3908106f92e1d43247a2fcfa33d639f0cda63667
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/acos_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_acos : public TensorIteratorBase {
+    
+    
+    void meta(const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_cpu_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4ca942ce0a005a46c4c49f7e3b793faaacc7b660
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_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> adaptive_max_pool3d(const at::Tensor & self, at::IntArrayRef output_size);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> adaptive_max_pool3d_out(at::Tensor & out, at::Tensor & indices, const at::Tensor & self, at::IntArrayRef output_size);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> adaptive_max_pool3d_outf(const at::Tensor & self, at::IntArrayRef output_size, at::Tensor & out, at::Tensor & indices);
+
+} // namespace cpu
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/bitwise_and_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/bitwise_and_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..15980402c596e8e08b9e62380958c4ed42c6397b
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/bitwise_and_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#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 bitwise_and(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & bitwise_and_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & bitwise_and_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & bitwise_and_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor bitwise_and(const at::Scalar & self, const at::Tensor & other);
+TORCH_API at::Tensor & bitwise_and_out(at::Tensor & out, const at::Scalar & self, const at::Tensor & other);
+TORCH_API at::Tensor & bitwise_and_outf(const at::Scalar & self, const at::Tensor & other, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/col_indices_copy_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/col_indices_copy_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..72111bbe0846ea0e7288a5336cb112884989d501
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/col_indices_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 col_indices_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::col_indices_copy")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "col_indices_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 col_indices_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::col_indices_copy")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "col_indices_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
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/copysign_cpu_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/copysign_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b42e95594833fecf54993c55754f14eced414d26
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/copysign_cpu_dispatch.h
@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API at::Tensor copysign(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & copysign_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & copysign_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & copysign_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace cpu
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/diagonal.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/diagonal.h
new file mode 100644
index 0000000000000000000000000000000000000000..7875db44d41030ade17e60d206a7de7243f5de7f
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/diagonal.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/diagonal_ops.h>
+
+namespace at {
+
+
+// aten::diagonal(Tensor(a) self, int offset=0, int dim1=0, int dim2=1) -> Tensor(a)
+inline at::Tensor diagonal(const at::Tensor & self, int64_t offset=0, int64_t dim1=0, int64_t dim2=1) {
+    return at::_ops::diagonal::call(self, offset, dim1, dim2);
+}
+
+// aten::diagonal.Dimname(Tensor(a) self, *, Dimname outdim, Dimname dim1, Dimname dim2, int offset=0) -> Tensor(a)
+inline at::Tensor diagonal(const at::Tensor & self, at::Dimname outdim, at::Dimname dim1, at::Dimname dim2, int64_t offset=0) {
+    return at::_ops::diagonal_Dimname::call(self, outdim, dim1, dim2, offset);
+}
+
+}
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/diagonal_backward_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/diagonal_backward_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6065622b0f1419b0ab221c4171b93e0bc1023ec3
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/diagonal_backward_compositeexplicitautograd_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 compositeexplicitautograd {
+
+TORCH_API at::Tensor diagonal_backward(const at::Tensor & grad_output, at::IntArrayRef input_sizes, int64_t offset, int64_t dim1, int64_t dim2);
+TORCH_API at::Tensor diagonal_backward_symint(const at::Tensor & grad_output, c10::SymIntArrayRef input_sizes, int64_t offset, int64_t dim1, int64_t dim2);
+TORCH_API at::Tensor & diagonal_backward_out(at::Tensor & out, const at::Tensor & grad_output, at::IntArrayRef input_sizes, int64_t offset, int64_t dim1, int64_t dim2);
+TORCH_API at::Tensor & diagonal_backward_outf(const at::Tensor & grad_output, at::IntArrayRef input_sizes, int64_t offset, int64_t dim1, int64_t dim2, at::Tensor & out);
+TORCH_API at::Tensor & diagonal_backward_symint_out(at::Tensor & out, const at::Tensor & grad_output, c10::SymIntArrayRef input_sizes, int64_t offset, int64_t dim1, int64_t dim2);
+TORCH_API at::Tensor & diagonal_backward_symint_outf(const at::Tensor & grad_output, c10::SymIntArrayRef input_sizes, int64_t offset, int64_t dim1, int64_t dim2, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/embedding_renorm_cpu_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/embedding_renorm_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..804c77bb81c066d11ae990e858b45fb8ae7f714b
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/embedding_renorm_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 & embedding_renorm_(at::Tensor & self, const at::Tensor & indices, double max_norm, double norm_type);
+
+} // namespace cpu
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/embedding_renorm_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/embedding_renorm_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..0cbe45d67ee1690d2539c3ef8a90124397b07f82
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/embedding_renorm_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 embedding_renorm(const at::Tensor & self, const at::Tensor & indices, double max_norm, double norm_type);
+TORCH_API at::Tensor & embedding_renorm_out(const at::Tensor & self, const at::Tensor & indices, double max_norm, double norm_type, at::Tensor & out);
+TORCH_API at::Tensor & embedding_renorm_cpu_(at::Tensor & self, const at::Tensor & indices, double max_norm, double norm_type);
+TORCH_API at::Tensor & embedding_renorm_cuda_(at::Tensor & self, const at::Tensor & indices, double max_norm, double norm_type);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_cachemask_backward.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_cachemask_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..cc214d98dcf278bf2ab43c9baa3757c8d5501c28
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_cachemask_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_per_tensor_affine_cachemask_backward_ops.h>
+
+namespace at {
+
+
+// aten::fake_quantize_per_tensor_affine_cachemask_backward(Tensor grad, Tensor mask) -> Tensor
+inline at::Tensor fake_quantize_per_tensor_affine_cachemask_backward(const at::Tensor & grad, const at::Tensor & mask) {
+    return at::_ops::fake_quantize_per_tensor_affine_cachemask_backward::call(grad, mask);
+}
+
+}
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fft_ifftn_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fft_ifftn_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..5f1ffc97b65216c00a158fd9ec5621d1bd580b73
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fft_ifftn_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 fft_ifftn_symint(const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::OptionalIntArrayRef dim=::std::nullopt, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_ifftn_symint_out(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::OptionalIntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fliplr_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fliplr_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..863d512967a0892eae10b78674052b3d856c32ec
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fliplr_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 fliplr(const at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fractional_max_pool3d_meta.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fractional_max_pool3d_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..667dce4fc27ff66bfa3a710a9f780a55ccc2a2cb
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fractional_max_pool3d_meta.h
@@ -0,0 +1,239 @@
+#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_fractional_max_pool3d : public at::impl::MetaBase {
+    
+                template <bool POOLSIZET = false, bool POOLSIZEH = false, bool POOLSIZEW = false, bool OUTPUTT = false, bool OUTPUTH = false, bool OUTPUTW = false, bool NUMBATCH = false, bool NUMPLANES = false, bool INPUTT = false, bool INPUTH = false, bool INPUTW = false>
+                struct TORCH_API precompute_out {
+                    
+                    precompute_out<true, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> set_poolSizeT(int64_t value) {
+                        static_assert(POOLSIZET == false, "poolSizeT already set");
+                        precompute_out<true, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> ret;
+ret.poolSizeT = value;
+ret.poolSizeH = this->poolSizeH;
+ret.poolSizeW = this->poolSizeW;
+ret.outputT = this->outputT;
+ret.outputH = this->outputH;
+ret.outputW = this->outputW;
+ret.numBatch = this->numBatch;
+ret.numPlanes = this->numPlanes;
+ret.inputT = this->inputT;
+ret.inputH = this->inputH;
+ret.inputW = this->inputW;
+return ret;
+                    }
+                
+
+                    precompute_out<POOLSIZET, true, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> set_poolSizeH(int64_t value) {
+                        static_assert(POOLSIZEH == false, "poolSizeH already set");
+                        precompute_out<POOLSIZET, true, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> ret;
+ret.poolSizeT = this->poolSizeT;
+ret.poolSizeH = value;
+ret.poolSizeW = this->poolSizeW;
+ret.outputT = this->outputT;
+ret.outputH = this->outputH;
+ret.outputW = this->outputW;
+ret.numBatch = this->numBatch;
+ret.numPlanes = this->numPlanes;
+ret.inputT = this->inputT;
+ret.inputH = this->inputH;
+ret.inputW = this->inputW;
+return ret;
+                    }
+                
+
+                    precompute_out<POOLSIZET, POOLSIZEH, true, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> set_poolSizeW(int64_t value) {
+                        static_assert(POOLSIZEW == false, "poolSizeW already set");
+                        precompute_out<POOLSIZET, POOLSIZEH, true, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> ret;
+ret.poolSizeT = this->poolSizeT;
+ret.poolSizeH = this->poolSizeH;
+ret.poolSizeW = value;
+ret.outputT = this->outputT;
+ret.outputH = this->outputH;
+ret.outputW = this->outputW;
+ret.numBatch = this->numBatch;
+ret.numPlanes = this->numPlanes;
+ret.inputT = this->inputT;
+ret.inputH = this->inputH;
+ret.inputW = this->inputW;
+return ret;
+                    }
+                
+
+                    precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, true, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> set_outputT(int64_t value) {
+                        static_assert(OUTPUTT == false, "outputT already set");
+                        precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, true, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> ret;
+ret.poolSizeT = this->poolSizeT;
+ret.poolSizeH = this->poolSizeH;
+ret.poolSizeW = this->poolSizeW;
+ret.outputT = value;
+ret.outputH = this->outputH;
+ret.outputW = this->outputW;
+ret.numBatch = this->numBatch;
+ret.numPlanes = this->numPlanes;
+ret.inputT = this->inputT;
+ret.inputH = this->inputH;
+ret.inputW = this->inputW;
+return ret;
+                    }
+                
+
+                    precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, true, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> set_outputH(int64_t value) {
+                        static_assert(OUTPUTH == false, "outputH already set");
+                        precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, true, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> ret;
+ret.poolSizeT = this->poolSizeT;
+ret.poolSizeH = this->poolSizeH;
+ret.poolSizeW = this->poolSizeW;
+ret.outputT = this->outputT;
+ret.outputH = value;
+ret.outputW = this->outputW;
+ret.numBatch = this->numBatch;
+ret.numPlanes = this->numPlanes;
+ret.inputT = this->inputT;
+ret.inputH = this->inputH;
+ret.inputW = this->inputW;
+return ret;
+                    }
+                
+
+                    precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, true, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> set_outputW(int64_t value) {
+                        static_assert(OUTPUTW == false, "outputW already set");
+                        precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, true, NUMBATCH, NUMPLANES, INPUTT, INPUTH, INPUTW> ret;
+ret.poolSizeT = this->poolSizeT;
+ret.poolSizeH = this->poolSizeH;
+ret.poolSizeW = this->poolSizeW;
+ret.outputT = this->outputT;
+ret.outputH = this->outputH;
+ret.outputW = value;
+ret.numBatch = this->numBatch;
+ret.numPlanes = this->numPlanes;
+ret.inputT = this->inputT;
+ret.inputH = this->inputH;
+ret.inputW = this->inputW;
+return ret;
+                    }
+                
+
+                    precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, true, NUMPLANES, INPUTT, INPUTH, INPUTW> set_numBatch(int64_t value) {
+                        static_assert(NUMBATCH == false, "numBatch already set");
+                        precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, true, NUMPLANES, INPUTT, INPUTH, INPUTW> ret;
+ret.poolSizeT = this->poolSizeT;
+ret.poolSizeH = this->poolSizeH;
+ret.poolSizeW = this->poolSizeW;
+ret.outputT = this->outputT;
+ret.outputH = this->outputH;
+ret.outputW = this->outputW;
+ret.numBatch = value;
+ret.numPlanes = this->numPlanes;
+ret.inputT = this->inputT;
+ret.inputH = this->inputH;
+ret.inputW = this->inputW;
+return ret;
+                    }
+                
+
+                    precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, true, INPUTT, INPUTH, INPUTW> set_numPlanes(int64_t value) {
+                        static_assert(NUMPLANES == false, "numPlanes already set");
+                        precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, true, INPUTT, INPUTH, INPUTW> ret;
+ret.poolSizeT = this->poolSizeT;
+ret.poolSizeH = this->poolSizeH;
+ret.poolSizeW = this->poolSizeW;
+ret.outputT = this->outputT;
+ret.outputH = this->outputH;
+ret.outputW = this->outputW;
+ret.numBatch = this->numBatch;
+ret.numPlanes = value;
+ret.inputT = this->inputT;
+ret.inputH = this->inputH;
+ret.inputW = this->inputW;
+return ret;
+                    }
+                
+
+                    precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, true, INPUTH, INPUTW> set_inputT(int64_t value) {
+                        static_assert(INPUTT == false, "inputT already set");
+                        precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, true, INPUTH, INPUTW> ret;
+ret.poolSizeT = this->poolSizeT;
+ret.poolSizeH = this->poolSizeH;
+ret.poolSizeW = this->poolSizeW;
+ret.outputT = this->outputT;
+ret.outputH = this->outputH;
+ret.outputW = this->outputW;
+ret.numBatch = this->numBatch;
+ret.numPlanes = this->numPlanes;
+ret.inputT = value;
+ret.inputH = this->inputH;
+ret.inputW = this->inputW;
+return ret;
+                    }
+                
+
+                    precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, true, INPUTW> set_inputH(int64_t value) {
+                        static_assert(INPUTH == false, "inputH already set");
+                        precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, true, INPUTW> ret;
+ret.poolSizeT = this->poolSizeT;
+ret.poolSizeH = this->poolSizeH;
+ret.poolSizeW = this->poolSizeW;
+ret.outputT = this->outputT;
+ret.outputH = this->outputH;
+ret.outputW = this->outputW;
+ret.numBatch = this->numBatch;
+ret.numPlanes = this->numPlanes;
+ret.inputT = this->inputT;
+ret.inputH = value;
+ret.inputW = this->inputW;
+return ret;
+                    }
+                
+
+                    precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, true> set_inputW(int64_t value) {
+                        static_assert(INPUTW == false, "inputW already set");
+                        precompute_out<POOLSIZET, POOLSIZEH, POOLSIZEW, OUTPUTT, OUTPUTH, OUTPUTW, NUMBATCH, NUMPLANES, INPUTT, INPUTH, true> ret;
+ret.poolSizeT = this->poolSizeT;
+ret.poolSizeH = this->poolSizeH;
+ret.poolSizeW = this->poolSizeW;
+ret.outputT = this->outputT;
+ret.outputH = this->outputH;
+ret.outputW = this->outputW;
+ret.numBatch = this->numBatch;
+ret.numPlanes = this->numPlanes;
+ret.inputT = this->inputT;
+ret.inputH = this->inputH;
+ret.inputW = value;
+return ret;
+                    }
+                
+                    int64_t poolSizeT;
+int64_t poolSizeH;
+int64_t poolSizeW;
+int64_t outputT;
+int64_t outputH;
+int64_t outputW;
+int64_t numBatch;
+int64_t numPlanes;
+int64_t inputT;
+int64_t inputH;
+int64_t inputW;
+            };
+    using meta_return_ty = precompute_out <true, true, true, true, true, true, true, true, true, true, true>;
+    meta_return_ty meta(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples);
+};
+
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/gelu_cpu_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/gelu_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3cdc3996fdb7468b985c7cde9f894433e4c7b616
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/gelu_cpu_dispatch.h
@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API at::Tensor gelu(const at::Tensor & self, c10::string_view approximate="none");
+TORCH_API at::Tensor & gelu_out(at::Tensor & out, const at::Tensor & self, c10::string_view approximate="none");
+TORCH_API at::Tensor & gelu_outf(const at::Tensor & self, c10::string_view approximate, at::Tensor & out);
+TORCH_API at::Tensor & gelu_(at::Tensor & self, c10::string_view approximate="none");
+
+} // namespace cpu
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/grid_sampler_2d_cpu_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/grid_sampler_2d_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d0a004abcf0d0559eea81053433c3dd147578455
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/grid_sampler_2d_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 grid_sampler_2d(const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners);
+
+} // namespace cpu
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/i0_meta_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/i0_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a4f3e136ebe261b908fc48dd390af63e641f922a
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/i0_meta_dispatch.h
@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API at::Tensor i0(const at::Tensor & self);
+TORCH_API at::Tensor & i0_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & i0_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & i0_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/index_fill_meta_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/index_fill_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..edf36d849dd34827357ff07afc5c00348c748de7
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/index_fill_meta_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 meta {
+
+TORCH_API at::Tensor & index_fill_(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value);
+TORCH_API at::Tensor & index_fill_(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & value);
+
+} // namespace meta
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/item.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/item.h
new file mode 100644
index 0000000000000000000000000000000000000000..40a7273d885d2f286bef65eaaf106ee756259482
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/item.h
@@ -0,0 +1,26 @@
+#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/item_ops.h>
+
+namespace at {
+
+
+
+}
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/lcm_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/lcm_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..ad560922ed827d4eeac753bd790fefe2f8b7ce94
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/lcm_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 lcm_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::lcm")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "lcm.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 lcm {
+  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::lcm")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "lcm(Tensor self, Tensor other) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & other);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other);
+};
+
+struct TORCH_API lcm_ {
+  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::lcm_")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "lcm_(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);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/leaky_relu_cpu_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/leaky_relu_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..20aa6e98c30c69ec0a4a49c77f72bf736683803b
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/leaky_relu_cpu_dispatch.h
@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API at::Tensor leaky_relu(const at::Tensor & self, const at::Scalar & negative_slope=0.01);
+TORCH_API at::Tensor & leaky_relu_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & negative_slope=0.01);
+TORCH_API at::Tensor & leaky_relu_outf(const at::Tensor & self, const at::Scalar & negative_slope, at::Tensor & out);
+TORCH_API at::Tensor & leaky_relu_(at::Tensor & self, const at::Scalar & negative_slope=0.01);
+
+} // namespace cpu
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_cholesky_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_cholesky_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..e0d09b256eca126d36739274cbe7ab093d2655d6
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_cholesky_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_cholesky(const at::Tensor & self, bool upper=false);
+TORCH_API at::Tensor & linalg_cholesky_out(const at::Tensor & self, bool upper, at::Tensor & out);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_svdvals_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_svdvals_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..36e944d69f148cca8781205645d18cc88a7252c3
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_svdvals_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_svdvals(const at::Tensor & A, ::std::optional<c10::string_view> driver=::std::nullopt);
+TORCH_API at::Tensor & linalg_svdvals_out(const at::Tensor & A, ::std::optional<c10::string_view> driver, at::Tensor & out);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/max_pool2d_with_indices_backward_cuda_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/max_pool2d_with_indices_backward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f90d59797e717c75d7e77f85dab5f0a2113a7df5
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/max_pool2d_with_indices_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 max_pool2d_with_indices_backward(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool ceil_mode, const at::Tensor & indices);
+TORCH_API at::Tensor & max_pool2d_with_indices_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, at::IntArrayRef dilation, bool ceil_mode, const at::Tensor & indices);
+TORCH_API at::Tensor & max_pool2d_with_indices_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool ceil_mode, const at::Tensor & indices, at::Tensor & grad_input);
+
+} // namespace cuda
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/miopen_convolution_add_relu_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/miopen_convolution_add_relu_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..b1e9ead9269d2960c316d6a69027a65c6de0840f
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/miopen_convolution_add_relu_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 miopen_convolution_add_relu {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, const ::std::optional<at::Scalar> &, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymInt);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::miopen_convolution_add_relu")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "miopen_convolution_add_relu(Tensor self, Tensor weight, Tensor z, Scalar? alpha, Tensor? bias, SymInt[] stride, SymInt[] padding, SymInt[] dilation, SymInt groups) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & weight, const at::Tensor & z, const ::std::optional<at::Scalar> & alpha, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, c10::SymInt groups);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & weight, const at::Tensor & z, const ::std::optional<at::Scalar> & alpha, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, c10::SymInt groups);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mish_cpu_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mish_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..806e108437c6b578d888da3c02051172513ce3f7
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mish_cpu_dispatch.h
@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API at::Tensor mish(const at::Tensor & self);
+TORCH_API at::Tensor & mish_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & mish_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & mish_(at::Tensor & self);
+
+} // namespace cpu
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mkldnn_max_pool3d_backward.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mkldnn_max_pool3d_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..d7d8b6d876748fdd92585f33372fdd77f040db1b
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mkldnn_max_pool3d_backward.h
@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/mkldnn_max_pool3d_backward_ops.h>
+
+namespace at {
+
+
+// aten::mkldnn_max_pool3d_backward(Tensor grad_output, Tensor output, Tensor input, int[3] kernel_size, int[3] stride=[], int[3] padding=0, int[3] dilation=1, bool ceil_mode=False) -> Tensor
+inline at::Tensor mkldnn_max_pool3d_backward(const at::Tensor & grad_output, const at::Tensor & output, const at::Tensor & input, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, at::IntArrayRef dilation=1, bool ceil_mode=false) {
+    return at::_ops::mkldnn_max_pool3d_backward::call(grad_output, output, input, kernel_size, stride, padding, dilation, ceil_mode);
+}
+
+// aten::mkldnn_max_pool3d_backward.out(Tensor grad_output, Tensor output, Tensor input, int[3] kernel_size, int[3] stride=[], int[3] padding=0, int[3] dilation=1, bool ceil_mode=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & mkldnn_max_pool3d_backward_out(at::Tensor & out, const at::Tensor & grad_output, const at::Tensor & output, const at::Tensor & input, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, at::IntArrayRef dilation=1, bool ceil_mode=false) {
+    return at::_ops::mkldnn_max_pool3d_backward_out::call(grad_output, output, input, kernel_size, stride, padding, dilation, ceil_mode, out);
+}
+// aten::mkldnn_max_pool3d_backward.out(Tensor grad_output, Tensor output, Tensor input, int[3] kernel_size, int[3] stride=[], int[3] padding=0, int[3] dilation=1, bool ceil_mode=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & mkldnn_max_pool3d_backward_outf(const at::Tensor & grad_output, const at::Tensor & output, const at::Tensor & input, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool ceil_mode, at::Tensor & out) {
+    return at::_ops::mkldnn_max_pool3d_backward_out::call(grad_output, output, input, kernel_size, stride, padding, dilation, ceil_mode, out);
+}
+
+}
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mv_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mv_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..fa0adc3aa2fb42b5e3cb71e4fa4f6e50065f04ad
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mv_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 mv {
+  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::mv")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "mv(Tensor self, Tensor vec) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & vec);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & vec);
+};
+
+struct TORCH_API mv_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::mv")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "mv.out(Tensor self, Tensor vec, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & vec, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & vec, at::Tensor & out);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/native_batch_norm_backward_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/native_batch_norm_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..5568c19acaed4407776fe11116a8e963bc125faf
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/native_batch_norm_backward_ops.h
@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API native_batch_norm_backward {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, bool, double, ::std::array<bool,3>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::native_batch_norm_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "native_batch_norm_backward(Tensor grad_out, Tensor input, Tensor? weight, Tensor? running_mean, Tensor? running_var, Tensor? save_mean, Tensor? save_invstd, bool train, float eps, bool[3] output_mask) -> (Tensor, Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> call(const at::Tensor & grad_out, const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, const ::std::optional<at::Tensor> & save_mean, const ::std::optional<at::Tensor> & save_invstd, bool train, double eps, ::std::array<bool,3> output_mask);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_out, const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, const ::std::optional<at::Tensor> & save_mean, const ::std::optional<at::Tensor> & save_invstd, bool train, double eps, ::std::array<bool,3> output_mask);
+};
+
+struct TORCH_API native_batch_norm_backward_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, bool, double, ::std::array<bool,3>, at::Tensor &, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::native_batch_norm_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "native_batch_norm_backward.out(Tensor grad_out, Tensor input, Tensor? weight, Tensor? running_mean, Tensor? running_var, Tensor? save_mean, Tensor? save_invstd, bool train, float eps, bool[3] output_mask, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))")
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> call(const at::Tensor & grad_out, const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, const ::std::optional<at::Tensor> & save_mean, const ::std::optional<at::Tensor> & save_invstd, bool train, double eps, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_out, const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, const ::std::optional<at::Tensor> & save_mean, const ::std::optional<at::Tensor> & save_invstd, bool train, double eps, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/native_batch_norm_cuda_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/native_batch_norm_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..84ddddb676649a67363e0f567d2831bee6c647b9
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/native_batch_norm_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,at::Tensor> native_batch_norm(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, bool training, double momentum, double eps);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_batch_norm_out(at::Tensor & out, at::Tensor & save_mean, at::Tensor & save_invstd, const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, bool training, double momentum, double eps);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_batch_norm_outf(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, bool training, double momentum, double eps, at::Tensor & out, at::Tensor & save_mean, at::Tensor & save_invstd);
+
+} // namespace cuda
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/permute.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/permute.h
new file mode 100644
index 0000000000000000000000000000000000000000..7524a584c32eefa730e7a48f7b56d9c7ab00651a
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/permute.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/permute_ops.h>
+
+namespace at {
+
+
+// aten::permute(Tensor(a) self, int[] dims) -> Tensor(a)
+inline at::Tensor permute(const at::Tensor & self, at::IntArrayRef dims) {
+    return at::_ops::permute::call(self, dims);
+}
+
+}
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/pixel_shuffle.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/pixel_shuffle.h
new file mode 100644
index 0000000000000000000000000000000000000000..82e85f94890219478e37e8f5429e3ef73d801c8e
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/pixel_shuffle.h
@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/pixel_shuffle_ops.h>
+
+namespace at {
+
+
+// aten::pixel_shuffle(Tensor self, int upscale_factor) -> Tensor
+inline at::Tensor pixel_shuffle(const at::Tensor & self, int64_t upscale_factor) {
+    return at::_ops::pixel_shuffle::call(self, upscale_factor);
+}
+
+// aten::pixel_shuffle.out(Tensor self, int upscale_factor, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & pixel_shuffle_out(at::Tensor & out, const at::Tensor & self, int64_t upscale_factor) {
+    return at::_ops::pixel_shuffle_out::call(self, upscale_factor, out);
+}
+// aten::pixel_shuffle.out(Tensor self, int upscale_factor, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & pixel_shuffle_outf(const at::Tensor & self, int64_t upscale_factor, at::Tensor & out) {
+    return at::_ops::pixel_shuffle_out::call(self, upscale_factor, out);
+}
+
+}
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/quantize_per_channel_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/quantize_per_channel_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..543b8ad39015f0cf7b18df4ae5b1d46205ed3e0b
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/quantize_per_channel_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 & quantize_per_channel_out(const at::Tensor & self, const at::Tensor & scales, const at::Tensor & zero_points, int64_t axis, at::ScalarType dtype, at::Tensor & out);
+TORCH_API at::Tensor quantize_per_channel(const at::Tensor & self, const at::Tensor & scales, const at::Tensor & zero_points, int64_t axis, at::ScalarType dtype);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/quantized_rnn_relu_cell_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/quantized_rnn_relu_cell_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..6672ec0739c14e88e7e7eb2cbf94c98dc6f0ef86
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/quantized_rnn_relu_cell_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 quantized_rnn_relu_cell {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Scalar &, const at::Scalar &, const at::Scalar &, 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::quantized_rnn_relu_cell")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "quantized_rnn_relu_cell(Tensor input, Tensor hx, Tensor w_ih, Tensor w_hh, Tensor b_ih, Tensor b_hh, Tensor packed_ih, Tensor packed_hh, Tensor col_offsets_ih, Tensor col_offsets_hh, Scalar scale_ih, Scalar scale_hh, Scalar zero_point_ih, Scalar zero_point_hh) -> Tensor")
+  static at::Tensor call(const at::Tensor & input, const at::Tensor & hx, const at::Tensor & w_ih, const at::Tensor & w_hh, const at::Tensor & b_ih, const at::Tensor & b_hh, const at::Tensor & packed_ih, const at::Tensor & packed_hh, const at::Tensor & col_offsets_ih, const at::Tensor & col_offsets_hh, const at::Scalar & scale_ih, const at::Scalar & scale_hh, const at::Scalar & zero_point_ih, const at::Scalar & zero_point_hh);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & hx, const at::Tensor & w_ih, const at::Tensor & w_hh, const at::Tensor & b_ih, const at::Tensor & b_hh, const at::Tensor & packed_ih, const at::Tensor & packed_hh, const at::Tensor & col_offsets_ih, const at::Tensor & col_offsets_hh, const at::Scalar & scale_ih, const at::Scalar & scale_hh, const at::Scalar & zero_point_ih, const at::Scalar & zero_point_hh);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/randperm_cuda_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/randperm_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..607e4571992f3dc4dac6ab40c52cce69d3e54a50
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/randperm_cuda_dispatch.h
@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor & randperm_out(at::Tensor & out, int64_t n, ::std::optional<at::Generator> generator);
+TORCH_API at::Tensor & randperm_outf(int64_t n, ::std::optional<at::Generator> generator, at::Tensor & out);
+TORCH_API at::Tensor & randperm_symint_out(at::Tensor & out, c10::SymInt n, ::std::optional<at::Generator> generator);
+TORCH_API at::Tensor & randperm_symint_outf(c10::SymInt n, ::std::optional<at::Generator> generator, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/remainder_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/remainder_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..590350e9cf15c4b92ab38b72ee1baaa3d61937e7
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/remainder_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>
+#include <ATen/ops/remainder_meta.h>
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor remainder(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & remainder_out(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & remainder_(at::Tensor & self, const at::Scalar & other);
+struct TORCH_API structured_remainder_out : public at::meta::structured_remainder_Tensor {
+void impl(const at::Tensor & self, const at::Tensor & other, const at::Tensor & out);
+};
+TORCH_API at::Tensor & remainder_Scalar_Tensor_out(const at::Scalar & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor remainder(const at::Scalar & self, const at::Tensor & other);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/repeat_interleave_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/repeat_interleave_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..ecaf1775357e1eb1313d6ce90a145957d29875ad
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/repeat_interleave_native.h
@@ -0,0 +1,25 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <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 & repeat_interleave_Tensor_out_symint(const at::Tensor & repeats, ::std::optional<c10::SymInt> output_size, at::Tensor & out);
+TORCH_API at::Tensor repeat_interleave_cpu(const at::Tensor & repeats, ::std::optional<int64_t> output_size=::std::nullopt);
+TORCH_API at::Tensor repeat_interleave_cuda(const at::Tensor & repeats, ::std::optional<int64_t> output_size=::std::nullopt);
+TORCH_API at::Tensor repeat_interleave_symint(const at::Tensor & self, const at::Tensor & repeats, ::std::optional<int64_t> dim=::std::nullopt, ::std::optional<c10::SymInt> output_size=::std::nullopt);
+TORCH_API at::Tensor repeat_interleave_symint(const at::Tensor & self, c10::SymInt repeats, ::std::optional<int64_t> dim=::std::nullopt, ::std::optional<c10::SymInt> output_size=::std::nullopt);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/result_type_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/result_type_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..a326db7a9ca7b4ecfff5d2e27112d30df7232033
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/result_type_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::ScalarType result_type(const at::Tensor & tensor, const at::Tensor & other);
+TORCH_API at::ScalarType result_type(const at::Tensor & tensor, const at::Scalar & other);
+TORCH_API at::ScalarType result_type(const at::Scalar & scalar, const at::Tensor & tensor);
+TORCH_API at::ScalarType result_type(const at::Scalar & scalar1, const at::Scalar & scalar2);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/retain_grad_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/retain_grad_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..f80f2722957eb25ac7f735dfa0e628e87a442e79
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/retain_grad_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 retain_grad {
+  using schema = void (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::retain_grad")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "retain_grad(Tensor(a!) self) -> ()")
+  static void call(at::Tensor & self);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/row_stack_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/row_stack_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..1c3aeeed16af08b94d2adba2be62cc7dc371dfaf
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/row_stack_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 row_stack(at::TensorList tensors);
+TORCH_API at::Tensor & row_stack_out(at::Tensor & out, at::TensorList tensors);
+TORCH_API at::Tensor & row_stack_outf(at::TensorList tensors, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/rrelu.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/rrelu.h
new file mode 100644
index 0000000000000000000000000000000000000000..ef9c7264588726e923b92aea6f2cb4b6e11ee1db
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/rrelu.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/rrelu_ops.h>
+
+namespace at {
+
+
+// aten::rrelu(Tensor self, Scalar lower=0.125, Scalar upper=0.3333333333333333, bool training=False, Generator? generator=None) -> Tensor
+inline at::Tensor rrelu(const at::Tensor & self, const at::Scalar & lower=0.125, const at::Scalar & upper=0.3333333333333333, bool training=false, ::std::optional<at::Generator> generator=::std::nullopt) {
+    return at::_ops::rrelu::call(self, lower, upper, training, generator);
+}
+
+// aten::rrelu_(Tensor(a!) self, Scalar lower=0.125, Scalar upper=0.3333333333333333, bool training=False, Generator? generator=None) -> Tensor(a!)
+inline at::Tensor & rrelu_(at::Tensor & self, const at::Scalar & lower=0.125, const at::Scalar & upper=0.3333333333333333, bool training=false, ::std::optional<at::Generator> generator=::std::nullopt) {
+    return at::_ops::rrelu_::call(self, lower, upper, training, generator);
+}
+
+}
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/rsqrt_cpu_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/rsqrt_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c8ca18bbb3faa59670c89914618871d56e53b668
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/rsqrt_cpu_dispatch.h
@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API at::Tensor rsqrt(const at::Tensor & self);
+TORCH_API at::Tensor & rsqrt_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & rsqrt_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & rsqrt_(at::Tensor & self);
+
+} // namespace cpu
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/silu_backward_meta.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/silu_backward_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..e059b4335d20b4cf6e2cff821a3ce3c0f0c820e4
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/silu_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_silu_backward : public TensorIteratorBase {
+    
+    
+    void meta(const at::Tensor & grad_output, const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/slice_scatter_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/slice_scatter_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..50dbbf7e82d1718e99068faee18b3d46504da01e
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/slice_scatter_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & slice_scatter_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & src, int64_t dim=0, ::std::optional<int64_t> start=::std::nullopt, ::std::optional<int64_t> end=::std::nullopt, int64_t step=1);
+TORCH_API at::Tensor & slice_scatter_outf(const at::Tensor & self, const at::Tensor & src, int64_t dim, ::std::optional<int64_t> start, ::std::optional<int64_t> end, int64_t step, at::Tensor & out);
+TORCH_API at::Tensor & slice_scatter_symint_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & src, int64_t dim=0, ::std::optional<c10::SymInt> start=::std::nullopt, ::std::optional<c10::SymInt> end=::std::nullopt, c10::SymInt step=1);
+TORCH_API at::Tensor & slice_scatter_symint_outf(const at::Tensor & self, const at::Tensor & src, int64_t dim, ::std::optional<c10::SymInt> start, ::std::optional<c10::SymInt> end, c10::SymInt step, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/sort_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/sort_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..0944e03bce1c7e7d8200c50d223fad7fffbcd06a
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/sort_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> sort(const at::Tensor & self, int64_t dim=-1, bool descending=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> sort_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, int64_t dim=-1, bool descending=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> sort_outf(const at::Tensor & self, int64_t dim, bool descending, at::Tensor & values, at::Tensor & indices);
+
+} // namespace compositeexplicitautograd
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..236562d05ecfabc6eb6696f9f0f642909fdae87d
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u_ops.h
@@ -0,0 +1,83 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API special_chebyshev_polynomial_u {
+  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::special_chebyshev_polynomial_u")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "special_chebyshev_polynomial_u(Tensor x, Tensor n) -> Tensor")
+  static at::Tensor call(const at::Tensor & x, const at::Tensor & n);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & x, const at::Tensor & n);
+};
+
+struct TORCH_API special_chebyshev_polynomial_u_x_scalar {
+  using schema = at::Tensor (const at::Scalar &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::special_chebyshev_polynomial_u")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "x_scalar")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "special_chebyshev_polynomial_u.x_scalar(Scalar x, Tensor n) -> Tensor")
+  static at::Tensor call(const at::Scalar & x, const at::Tensor & n);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & x, const at::Tensor & n);
+};
+
+struct TORCH_API special_chebyshev_polynomial_u_n_scalar {
+  using schema = at::Tensor (const at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::special_chebyshev_polynomial_u")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "n_scalar")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "special_chebyshev_polynomial_u.n_scalar(Tensor x, Scalar n) -> Tensor")
+  static at::Tensor call(const at::Tensor & x, const at::Scalar & n);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & x, const at::Scalar & n);
+};
+
+struct TORCH_API special_chebyshev_polynomial_u_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::special_chebyshev_polynomial_u")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "special_chebyshev_polynomial_u.out(Tensor x, Tensor n, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & x, const at::Tensor & n, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & x, const at::Tensor & n, at::Tensor & out);
+};
+
+struct TORCH_API special_chebyshev_polynomial_u_x_scalar_out {
+  using schema = at::Tensor & (const at::Scalar &, 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::special_chebyshev_polynomial_u")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "x_scalar_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "special_chebyshev_polynomial_u.x_scalar_out(Scalar x, Tensor n, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Scalar & x, const at::Tensor & n, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & x, const at::Tensor & n, at::Tensor & out);
+};
+
+struct TORCH_API special_chebyshev_polynomial_u_n_scalar_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::special_chebyshev_polynomial_u")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "n_scalar_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "special_chebyshev_polynomial_u.n_scalar_out(Tensor x, Scalar n, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & x, const at::Scalar & n, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & x, const at::Scalar & n, at::Tensor & out);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..ec91712eb2135cc23d269551f94d06d3ae76d824
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_native.h
@@ -0,0 +1,27 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+#include <ATen/ops/special_laguerre_polynomial_l_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_special_laguerre_polynomial_l_out : public at::meta::structured_special_laguerre_polynomial_l {
+void impl(const at::Tensor & x, const at::Tensor & n, const at::Tensor & out);
+};
+TORCH_API at::Tensor special_laguerre_polynomial_l(const at::Scalar & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_laguerre_polynomial_l_out(const at::Scalar & x, const at::Tensor & n, at::Tensor & out);
+TORCH_API at::Tensor special_laguerre_polynomial_l(const at::Tensor & x, const at::Scalar & n);
+TORCH_API at::Tensor & special_laguerre_polynomial_l_out(const at::Tensor & x, const at::Scalar & n, at::Tensor & out);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_modified_bessel_i1_cpu_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_modified_bessel_i1_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..89a3420e6e8d57a3e2c8c35d810fccb7e63118ee
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_modified_bessel_i1_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 special_modified_bessel_i1(const at::Tensor & self);
+TORCH_API at::Tensor & special_modified_bessel_i1_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & special_modified_bessel_i1_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_w_meta_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_w_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d54a9b56b4808047748acc5b0580f931ff794e0a
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_w_meta_dispatch.h
@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API at::Tensor special_shifted_chebyshev_polynomial_w(const at::Tensor & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_shifted_chebyshev_polynomial_w_out(at::Tensor & out, const at::Tensor & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_shifted_chebyshev_polynomial_w_outf(const at::Tensor & x, const at::Tensor & n, at::Tensor & out);
+
+} // namespace meta
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/t.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/t.h
new file mode 100644
index 0000000000000000000000000000000000000000..05c4d9d5b802d7d6371bde31f3973a6eb29eda61
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/t.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/t_ops.h>
+
+namespace at {
+
+
+// aten::t(Tensor(a) self) -> Tensor(a)
+inline at::Tensor t(const at::Tensor & self) {
+    return at::_ops::t::call(self);
+}
+
+}
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/threshold_backward_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/threshold_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..6430b8023e05eae06ddd1aa38c95c3ee715d87c1
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/threshold_backward_native.h
@@ -0,0 +1,29 @@
+#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>
+#include <ATen/ops/threshold_backward_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_threshold_backward_out : public at::meta::structured_threshold_backward {
+void impl(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & threshold, const at::Tensor & grad_input);
+};
+TORCH_API at::Tensor threshold_backwards_nested(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & threshold);
+TORCH_API at::Tensor threshold_backward_sparse(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & threshold);
+TORCH_API at::Tensor & threshold_backward_sparse_out(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & threshold, at::Tensor & grad_input);
+TORCH_API at::Tensor threshold_backward_sparse_compressed(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & threshold);
+TORCH_API at::Tensor & threshold_backward_sparse_compressed_out(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & threshold, at::Tensor & grad_input);
+TORCH_API at::Tensor mkldnn_relu_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & threshold);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/to_mkldnn_backward_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/to_mkldnn_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..20f625ad777df606cda3af92e7b3896d8c3c7a61
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/to_mkldnn_backward_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 to_mkldnn_backward {
+  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::to_mkldnn_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "to_mkldnn_backward(Tensor grad, Tensor input) -> Tensor")
+  static at::Tensor call(const at::Tensor & grad, const at::Tensor & input);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad, const at::Tensor & input);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/upsample_bicubic2d_backward.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/upsample_bicubic2d_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..1ef50aabdea2899ad12be535f073d23db1f50cd9
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/upsample_bicubic2d_backward.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/upsample_bicubic2d_backward_ops.h>
+
+namespace at {
+
+
+// aten::upsample_bicubic2d_backward.grad_input(Tensor grad_output, SymInt[2] output_size, SymInt[4] input_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) grad_input) -> Tensor(a!)
+inline at::Tensor & upsample_bicubic2d_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_backward_grad_input::call(grad_output, c10::fromIntArrayRefSlow(output_size), c10::fromIntArrayRefSlow(input_size), align_corners, scales_h, scales_w, grad_input);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & upsample_bicubic2d_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_backward_grad_input::call(grad_output, c10::fromIntArrayRefSlow(output_size), c10::fromIntArrayRefSlow(input_size), align_corners, scales_h, scales_w, grad_input);
+  }
+}
+
+// aten::upsample_bicubic2d_backward.grad_input(Tensor grad_output, SymInt[2] output_size, SymInt[4] input_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) grad_input) -> Tensor(a!)
+inline at::Tensor & upsample_bicubic2d_backward_outf(const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & grad_input) {
+    return at::_ops::upsample_bicubic2d_backward_grad_input::call(grad_output, c10::fromIntArrayRefSlow(output_size), c10::fromIntArrayRefSlow(input_size), align_corners, scales_h, scales_w, grad_input);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & upsample_bicubic2d_backward_outf(const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & grad_input) {
+    return at::_ops::upsample_bicubic2d_backward_grad_input::call(grad_output, c10::fromIntArrayRefSlow(output_size), c10::fromIntArrayRefSlow(input_size), align_corners, scales_h, scales_w, grad_input);
+  }
+}
+
+// aten::upsample_bicubic2d_backward.grad_input(Tensor grad_output, SymInt[2] output_size, SymInt[4] input_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) grad_input) -> Tensor(a!)
+inline at::Tensor & upsample_bicubic2d_backward_symint_out(at::Tensor & grad_input, const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_backward_grad_input::call(grad_output, output_size, input_size, align_corners, scales_h, scales_w, grad_input);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & upsample_bicubic2d_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_backward_grad_input::call(grad_output, output_size, input_size, align_corners, scales_h, scales_w, grad_input);
+  }
+}
+
+// aten::upsample_bicubic2d_backward.grad_input(Tensor grad_output, SymInt[2] output_size, SymInt[4] input_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) grad_input) -> Tensor(a!)
+inline at::Tensor & upsample_bicubic2d_backward_symint_outf(const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & grad_input) {
+    return at::_ops::upsample_bicubic2d_backward_grad_input::call(grad_output, output_size, input_size, align_corners, scales_h, scales_w, grad_input);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & upsample_bicubic2d_backward_outf(const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & grad_input) {
+    return at::_ops::upsample_bicubic2d_backward_grad_input::call(grad_output, output_size, input_size, align_corners, scales_h, scales_w, grad_input);
+  }
+}
+
+// aten::upsample_bicubic2d_backward(Tensor grad_output, SymInt[2] output_size, SymInt[4] input_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor
+inline at::Tensor upsample_bicubic2d_backward(const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_backward::call(grad_output, c10::fromIntArrayRefSlow(output_size), c10::fromIntArrayRefSlow(input_size), align_corners, scales_h, scales_w);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor upsample_bicubic2d_backward(const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_backward::call(grad_output, c10::fromIntArrayRefSlow(output_size), c10::fromIntArrayRefSlow(input_size), align_corners, scales_h, scales_w);
+  }
+}
+
+// aten::upsample_bicubic2d_backward(Tensor grad_output, SymInt[2] output_size, SymInt[4] input_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor
+inline at::Tensor upsample_bicubic2d_backward_symint(const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_backward::call(grad_output, output_size, input_size, align_corners, scales_h, scales_w);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor upsample_bicubic2d_backward(const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_backward::call(grad_output, output_size, input_size, align_corners, scales_h, scales_w);
+  }
+}
+
+}
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/upsample_trilinear3d_backward_cpu_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/upsample_trilinear3d_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4da0ac52555529307131ca16ac0314ce04b04f73
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/upsample_trilinear3d_backward_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API at::Tensor upsample_trilinear3d_backward(const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, bool align_corners, ::std::optional<double> scales_d=::std::nullopt, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor upsample_trilinear3d_backward_symint(const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, bool align_corners, ::std::optional<double> scales_d=::std::nullopt, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & upsample_trilinear3d_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, bool align_corners, ::std::optional<double> scales_d=::std::nullopt, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & upsample_trilinear3d_backward_outf(const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, bool align_corners, ::std::optional<double> scales_d, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & grad_input);
+TORCH_API at::Tensor & upsample_trilinear3d_backward_symint_out(at::Tensor & grad_input, const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, bool align_corners, ::std::optional<double> scales_d=::std::nullopt, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & upsample_trilinear3d_backward_symint_outf(const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, bool align_corners, ::std::optional<double> scales_d, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/vander_native.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/vander_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..96255830849d21a5c14b1651cdf37d4ff44a52ca
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/vander_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 vander(const at::Tensor & x, ::std::optional<int64_t> N=::std::nullopt, bool increasing=false);
+} // namespace native
+} // namespace at
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/vander_ops.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/vander_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..f5703cfd987f7b582fae3771b0e8e3e533e11983
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/vander_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 vander {
+  using schema = at::Tensor (const at::Tensor &, ::std::optional<int64_t>, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::vander")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "vander(Tensor x, int? N=None, bool increasing=False) -> Tensor")
+  static at::Tensor call(const at::Tensor & x, ::std::optional<int64_t> N, bool increasing);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & x, ::std::optional<int64_t> N, bool increasing);
+};
+
+}} // namespace at::_ops
diff --git a/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/view_copy_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/view_copy_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f2dd2c20b76e29daa972f8c5a6ecac15ff704e47
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/view_copy_compositeexplicitautogradnonfunctional_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 compositeexplicitautogradnonfunctional {
+
+TORCH_API at::Tensor view_copy(const at::Tensor & self, at::IntArrayRef size);
+TORCH_API at::Tensor view_copy_symint(const at::Tensor & self, c10::SymIntArrayRef size);
+TORCH_API at::Tensor view_copy(const at::Tensor & self, at::ScalarType dtype);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at