diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_adaptive_avg_pool2d_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_adaptive_avg_pool2d_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d6ea7d2ce35c016b8e34711d773557b43c5959c3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_adaptive_avg_pool2d_backward_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _adaptive_avg_pool2d_backward(const at::Tensor & grad_output, const at::Tensor & self);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_adaptive_avg_pool2d_backward_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_adaptive_avg_pool2d_backward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..48f9e6a94bb6fd272c4e91ccc5f8a348fb6cd8ed
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_adaptive_avg_pool2d_backward_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor _adaptive_avg_pool2d_backward(const at::Tensor & grad_output, const at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_adaptive_avg_pool3d_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_adaptive_avg_pool3d_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..4c03d60490dece89a9421f8f9201676a6522c936
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_adaptive_avg_pool3d_backward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _adaptive_avg_pool3d_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 const char* name = "aten::_adaptive_avg_pool3d_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_adaptive_avg_pool3d_backward(Tensor grad_output, Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad_output, const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self);
+};
+
+struct TORCH_API _adaptive_avg_pool3d_backward_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 const char* name = "aten::_adaptive_avg_pool3d_backward";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_adaptive_avg_pool3d_backward.out(Tensor grad_output, Tensor self, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & grad_output, const at::Tensor & self, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_add_batch_dim_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_add_batch_dim_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..ce3d4cc91753e37a33b2fb8e09b96b76e13d1da2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_add_batch_dim_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _add_batch_dim {
+  using schema = at::Tensor (const at::Tensor &, int64_t, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_add_batch_dim";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_add_batch_dim(Tensor self, int batch_dim, int level) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t batch_dim, int64_t level);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t batch_dim, int64_t level);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_addmm_activation_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_addmm_activation_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..50d7cce05d2a82eefae74722676792a760103dc7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_addmm_activation_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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__addmm_activation : public at::impl::MetaBase {
+
+
+    void meta(const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta, const at::Scalar & alpha, bool use_gelu);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_aminmax_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_aminmax_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..32ec31050bcb217753208d1604ea0e859d6071b3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_aminmax_ops.h
@@ -0,0 +1,67 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _aminmax {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_aminmax";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_aminmax(Tensor self) -> (Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+struct TORCH_API _aminmax_dim {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, int64_t, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_aminmax";
+  static constexpr const char* overload_name = "dim";
+  static constexpr const char* schema_str = "_aminmax.dim(Tensor self, int dim, bool keepdim=False) -> (Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, int64_t dim, bool keepdim);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, bool keepdim);
+};
+
+struct TORCH_API _aminmax_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_aminmax";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_aminmax.out(Tensor self, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, at::Tensor & out0, at::Tensor & out1);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Tensor & out0, at::Tensor & out1);
+};
+
+struct TORCH_API _aminmax_dim_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, int64_t, bool, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_aminmax";
+  static constexpr const char* overload_name = "dim_out";
+  static constexpr const char* schema_str = "_aminmax.dim_out(Tensor self, int dim, bool keepdim=False, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, int64_t dim, bool keepdim, at::Tensor & out0, at::Tensor & out1);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, bool keepdim, at::Tensor & out0, at::Tensor & out1);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_amp_foreach_non_finite_check_and_unscale_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_amp_foreach_non_finite_check_and_unscale_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4ac76806ceb02bc651e9c8f83c00bacd615a60c9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_amp_foreach_non_finite_check_and_unscale_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API void _amp_foreach_non_finite_check_and_unscale_(at::TensorList self, at::Tensor & found_inf, const at::Tensor & inv_scale);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_amp_foreach_non_finite_check_and_unscale_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_amp_foreach_non_finite_check_and_unscale_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..4b41d66c6000d24144bcf326f2d3e9a440714ced
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_amp_foreach_non_finite_check_and_unscale_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<::std::vector<at::Tensor>,at::Tensor> _amp_foreach_non_finite_check_and_unscale(at::TensorList self, const at::Tensor & found_inf, const at::Tensor & inv_scale);
+TORCH_API void _amp_foreach_non_finite_check_and_unscale_out(at::TensorList self, at::Tensor & found_inf, const at::Tensor & inv_scale, at::TensorList out);
+TORCH_API void _amp_foreach_non_finite_check_and_unscale_cpu_(at::TensorList self, at::Tensor & found_inf, const at::Tensor & inv_scale);
+TORCH_API void _amp_foreach_non_finite_check_and_unscale_cuda_(at::TensorList self, at::Tensor & found_inf, const at::Tensor & inv_scale);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_amp_update_scale_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_amp_update_scale_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..36315845e368dbeabf3af7efbea740a534e5a19c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_amp_update_scale_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _amp_update_scale_ {
+  using schema = at::Tensor & (at::Tensor &, at::Tensor &, const at::Tensor &, double, double, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_amp_update_scale_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_amp_update_scale_(Tensor(a!) self, Tensor(b!) growth_tracker, Tensor found_inf, float scale_growth_factor, float scale_backoff_factor, int growth_interval) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, at::Tensor & growth_tracker, const at::Tensor & found_inf, double scale_growth_factor, double scale_backoff_factor, int64_t growth_interval);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, at::Tensor & growth_tracker, const at::Tensor & found_inf, double scale_growth_factor, double scale_backoff_factor, int64_t growth_interval);
+};
+
+struct TORCH_API _amp_update_scale_out {
+  using schema = at::Tensor & (const at::Tensor &, at::Tensor &, const at::Tensor &, double, double, int64_t, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_amp_update_scale";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_amp_update_scale.out(Tensor self, Tensor(b!) growth_tracker, Tensor found_inf, float scale_growth_factor, float scale_backoff_factor, int growth_interval, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, at::Tensor & growth_tracker, const at::Tensor & found_inf, double scale_growth_factor, double scale_backoff_factor, int64_t growth_interval, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Tensor & growth_tracker, const at::Tensor & found_inf, double scale_growth_factor, double scale_backoff_factor, int64_t growth_interval, at::Tensor & out);
+};
+
+struct TORCH_API _amp_update_scale {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, double, double, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_amp_update_scale";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_amp_update_scale(Tensor self, Tensor growth_tracker, Tensor found_inf, float scale_growth_factor, float scale_backoff_factor, int growth_interval) -> (Tensor, Tensor growth_tracker_out)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, const at::Tensor & growth_tracker, const at::Tensor & found_inf, double scale_growth_factor, double scale_backoff_factor, int64_t growth_interval);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & growth_tracker, const at::Tensor & found_inf, double scale_growth_factor, double scale_backoff_factor, int64_t growth_interval);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_async_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_async_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..226e4a3efcf0b73534e03921bb143046692963ed
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_async_cpu_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API void _assert_async(const at::Tensor & self);
+TORCH_API void _assert_async(const at::Tensor & self, c10::string_view assert_msg);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_async_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_async_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..5ee01c15656b13bd917ec0a1c8e8e0dfdd6cedcc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_async_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void _assert_async_cpu(const at::Tensor & self);
+TORCH_API void _assert_async_cuda(const at::Tensor & self);
+TORCH_API void _assert_async_msg_cpu(const at::Tensor & self, c10::string_view assert_msg);
+TORCH_API void _assert_async_msg_cuda(const at::Tensor & self, c10::string_view assert_msg);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_scalar_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_scalar_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b44313227a8965c0723cd6dbe1724deaaf0eeed4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_scalar_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void _assert_scalar(const at::Scalar & self, c10::string_view assert_msg);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_tensor_metadata.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_tensor_metadata.h
new file mode 100644
index 0000000000000000000000000000000000000000..1275376b264493172e6280e276a3ccdbbfc71a48
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_assert_tensor_metadata.h
@@ -0,0 +1,53 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_assert_tensor_metadata_ops.h>
+
+namespace at {
+
+
+// aten::_assert_tensor_metadata(Tensor a, SymInt[]? size=None, SymInt[]? stride=None, ScalarType? dtype=None, *, Device? device=None, Layout? layout=None) -> ()
+inline void _assert_tensor_metadata(const at::Tensor & a, at::OptionalIntArrayRef size=::std::nullopt, at::OptionalIntArrayRef stride=::std::nullopt, ::std::optional<at::ScalarType> dtype=::std::nullopt, ::std::optional<at::Device> device=::std::nullopt, ::std::optional<at::Layout> layout=::std::nullopt) {
+    return at::_ops::_assert_tensor_metadata::call(a, size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*size)) : ::std::nullopt, stride.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*stride)) : ::std::nullopt, dtype, device, layout);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  void _assert_tensor_metadata(const at::Tensor & a, at::OptionalIntArrayRef size=::std::nullopt, at::OptionalIntArrayRef stride=::std::nullopt, ::std::optional<at::ScalarType> dtype=::std::nullopt, ::std::optional<at::Device> device=::std::nullopt, ::std::optional<at::Layout> layout=::std::nullopt) {
+    return at::_ops::_assert_tensor_metadata::call(a, size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*size)) : ::std::nullopt, stride.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*stride)) : ::std::nullopt, dtype, device, layout);
+  }
+}
+
+// aten::_assert_tensor_metadata(Tensor a, SymInt[]? size=None, SymInt[]? stride=None, ScalarType? dtype=None, *, Device? device=None, Layout? layout=None) -> ()
+inline void _assert_tensor_metadata_symint(const at::Tensor & a, at::OptionalSymIntArrayRef size=::std::nullopt, at::OptionalSymIntArrayRef stride=::std::nullopt, ::std::optional<at::ScalarType> dtype=::std::nullopt, ::std::optional<at::Device> device=::std::nullopt, ::std::optional<at::Layout> layout=::std::nullopt) {
+    return at::_ops::_assert_tensor_metadata::call(a, size, stride, dtype, device, layout);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  void _assert_tensor_metadata(const at::Tensor & a, at::OptionalSymIntArrayRef size=::std::nullopt, at::OptionalSymIntArrayRef stride=::std::nullopt, ::std::optional<at::ScalarType> dtype=::std::nullopt, ::std::optional<at::Device> device=::std::nullopt, ::std::optional<at::Layout> layout=::std::nullopt) {
+    return at::_ops::_assert_tensor_metadata::call(a, size, stride, dtype, device, layout);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_autocast_to_full_precision_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_autocast_to_full_precision_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..2613a66428f3cb1c9791917807dace7296331f09
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_autocast_to_full_precision_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _autocast_to_full_precision(const at::Tensor & self, bool cuda_enabled, bool cpu_enabled);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_backward.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..283436a91eb5ee7e7dc6d5b3aa3b55a9ec44027b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_backward.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_backward_ops.h>
+
+namespace at {
+
+
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..e0bddc3ddebc765890a264fabce60e80dd704f8d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_backward_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void _backward(const at::Tensor & self, at::TensorList inputs, const ::std::optional<at::Tensor> & gradient={}, ::std::optional<bool> retain_graph=::std::nullopt, bool create_graph=false);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_batch_norm_no_update.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_batch_norm_no_update.h
new file mode 100644
index 0000000000000000000000000000000000000000..caf169fc868908bf7f24f3ee6818f11e105af58c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_batch_norm_no_update.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_batch_norm_no_update_ops.h>
+
+namespace at {
+
+
+// aten::_batch_norm_no_update(Tensor input, Tensor? weight, Tensor? bias, Tensor? running_mean, Tensor? running_var, float momentum, float eps) -> (Tensor, Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> _batch_norm_no_update(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, double momentum, double eps) {
+    return at::_ops::_batch_norm_no_update::call(input, weight, bias, running_mean, running_var, momentum, eps);
+}
+
+// aten::_batch_norm_no_update.out(Tensor input, Tensor? weight, Tensor? bias, Tensor? running_mean, Tensor? running_var, float momentum, float eps, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2, Tensor(d!) out3) -> (Tensor(a!), Tensor(b!), Tensor(c!), Tensor(d!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> _batch_norm_no_update_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3, 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, double momentum, double eps) {
+    return at::_ops::_batch_norm_no_update_out::call(input, weight, bias, running_mean, running_var, momentum, eps, out0, out1, out2, out3);
+}
+// aten::_batch_norm_no_update.out(Tensor input, Tensor? weight, Tensor? bias, Tensor? running_mean, Tensor? running_var, float momentum, float eps, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2, Tensor(d!) out3) -> (Tensor(a!), Tensor(b!), Tensor(c!), Tensor(d!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> _batch_norm_no_update_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, double momentum, double eps, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3) {
+    return at::_ops::_batch_norm_no_update_out::call(input, weight, bias, running_mean, running_var, momentum, eps, out0, out1, out2, out3);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cast_Double_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cast_Double_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..29bb00d2ea76381c12dd04fd17c65bb999baddbd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cast_Double_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _cast_Double(const at::Tensor & self, bool non_blocking=false);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cast_Double_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cast_Double_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..f51b5beca2abe8a6b0f7d4c74c2833debd153ece
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cast_Double_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_Double {
+  using schema = at::Tensor (const at::Tensor &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_cast_Double";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_cast_Double(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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cast_Long_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cast_Long_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..20e651e26b61bdbd748a81715976eba2107767dc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cast_Long_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _cast_Long(const at::Tensor & self, bool non_blocking=false);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cdist_backward_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cdist_backward_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6fcb4d86e34cfce14f55a01b8899877832a4cc86
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cdist_backward_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _cdist_backward_out(at::Tensor & out, const at::Tensor & grad, const at::Tensor & x1, const at::Tensor & x2, double p, const at::Tensor & cdist);
+TORCH_API at::Tensor & _cdist_backward_outf(const at::Tensor & grad, const at::Tensor & x1, const at::Tensor & x2, double p, const at::Tensor & cdist, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cdist_backward_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cdist_backward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..2f65823d248c59c4776d461ee3200ff996998e50
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cdist_backward_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _cdist_backward(const at::Tensor & grad, const at::Tensor & x1, const at::Tensor & x2, double p, const at::Tensor & cdist);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cdist_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cdist_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..f9b80465aabf84f59fe9dd9074cfe1c0eaa1011f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cdist_backward_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _cdist_backward_out(const at::Tensor & grad, const at::Tensor & x1, const at::Tensor & x2, double p, const at::Tensor & cdist, at::Tensor & out);
+TORCH_API at::Tensor _cdist_backward(const at::Tensor & grad, const at::Tensor & x1, const at::Tensor & x2, double p, const at::Tensor & cdist);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cholesky_solve_helper_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cholesky_solve_helper_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..ea3c227c8288e8f9ec0388b1e4042c839dd760e9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cholesky_solve_helper_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _cholesky_solve_helper {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_cholesky_solve_helper";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_cholesky_solve_helper(Tensor self, Tensor A, bool upper) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & A, bool upper);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & A, bool upper);
+};
+
+struct TORCH_API _cholesky_solve_helper_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_cholesky_solve_helper";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_cholesky_solve_helper.out(Tensor self, Tensor A, bool upper, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & A, bool upper, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & A, bool upper, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_choose_qparams_per_tensor.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_choose_qparams_per_tensor.h
new file mode 100644
index 0000000000000000000000000000000000000000..e13feb2ca7d02c0d54a2a7d4c94a4a579809a4a8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_choose_qparams_per_tensor.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_choose_qparams_per_tensor_ops.h>
+
+namespace at {
+
+
+// aten::_choose_qparams_per_tensor(Tensor self, bool reduce_range=False) -> (float, int)
+inline ::std::tuple<double,int64_t> _choose_qparams_per_tensor(const at::Tensor & self, bool reduce_range=false) {
+    return at::_ops::_choose_qparams_per_tensor::call(self, reduce_range);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_chunk_cat.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_chunk_cat.h
new file mode 100644
index 0000000000000000000000000000000000000000..a3140cc8459c61d83371850f53628037cc61c5ba
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_chunk_cat.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_chunk_cat_ops.h>
+
+namespace at {
+
+
+// aten::_chunk_cat(Tensor[] tensors, int dim, int num_chunks) -> Tensor
+inline at::Tensor _chunk_cat(at::TensorList tensors, int64_t dim, int64_t num_chunks) {
+    return at::_ops::_chunk_cat::call(tensors, dim, num_chunks);
+}
+
+// aten::_chunk_cat.out(Tensor[] tensors, int dim, int num_chunks, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _chunk_cat_out(at::Tensor & out, at::TensorList tensors, int64_t dim, int64_t num_chunks) {
+    return at::_ops::_chunk_cat_out::call(tensors, dim, num_chunks, out);
+}
+// aten::_chunk_cat.out(Tensor[] tensors, int dim, int num_chunks, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _chunk_cat_outf(at::TensorList tensors, int64_t dim, int64_t num_chunks, at::Tensor & out) {
+    return at::_ops::_chunk_cat_out::call(tensors, dim, num_chunks, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_chunk_cat_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_chunk_cat_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..0c9fb1280e062604f43f16a8b57f1bec5f205905
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_chunk_cat_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _chunk_cat(at::TensorList tensors, int64_t dim, int64_t num_chunks);
+TORCH_API at::Tensor & _chunk_cat_out(at::Tensor & out, at::TensorList tensors, int64_t dim, int64_t num_chunks);
+TORCH_API at::Tensor & _chunk_cat_outf(at::TensorList tensors, int64_t dim, int64_t num_chunks, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_chunk_cat_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_chunk_cat_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..1ff19b1ef25984e32b9734707b8231ec392682d2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_chunk_cat_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _chunk_cat {
+  using schema = at::Tensor (at::TensorList, int64_t, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_chunk_cat";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_chunk_cat(Tensor[] tensors, int dim, int num_chunks) -> Tensor";
+  static at::Tensor call(at::TensorList tensors, int64_t dim, int64_t num_chunks);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList tensors, int64_t dim, int64_t num_chunks);
+};
+
+struct TORCH_API _chunk_cat_out {
+  using schema = at::Tensor & (at::TensorList, int64_t, int64_t, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_chunk_cat";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_chunk_cat.out(Tensor[] tensors, int dim, int num_chunks, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(at::TensorList tensors, int64_t dim, int64_t num_chunks, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList tensors, int64_t dim, int64_t num_chunks, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_coalesce_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_coalesce_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..a902e96a98aeeb1449836cc711395aa231bfb751
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_coalesce_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 const char* name = "aten::_coalesce";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* 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 const char* name = "aten::_coalesce";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* 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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_coalesced_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_coalesced_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f53d622d0886de8189a7a5589edabafff0854207
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_coalesced_meta_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _coalesced_(at::Tensor & self, bool coalesced);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_compute_linear_combination_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_compute_linear_combination_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..743a32b9d906763fe497057a3a0816fd2357ac4e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_compute_linear_combination_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _compute_linear_combination(const at::Tensor & input, const at::Tensor & coefficients);
+TORCH_API at::Tensor & _compute_linear_combination_out(const at::Tensor & input, const at::Tensor & coefficients, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_conj_copy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_conj_copy.h
new file mode 100644
index 0000000000000000000000000000000000000000..1bc07bae430712912168a5dcb29b3c306ae494e3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_conj_copy.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_conj_copy_ops.h>
+
+namespace at {
+
+
+// aten::_conj_copy(Tensor self) -> Tensor
+inline at::Tensor _conj_copy(const at::Tensor & self) {
+    return at::_ops::_conj_copy::call(self);
+}
+
+// aten::_conj_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _conj_copy_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::_conj_copy_out::call(self, out);
+}
+// aten::_conj_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _conj_copy_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::_conj_copy_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_indices_from_coo_to_csr_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_indices_from_coo_to_csr_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..95e25a18a4e199a7bbd096024b279fd09ef66196
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_indices_from_coo_to_csr_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _convert_indices_from_coo_to_csr(const at::Tensor & self, int64_t size, bool out_int32=false);
+TORCH_API at::Tensor & _convert_indices_from_coo_to_csr_out(at::Tensor & out, const at::Tensor & self, int64_t size, bool out_int32=false);
+TORCH_API at::Tensor & _convert_indices_from_coo_to_csr_outf(const at::Tensor & self, int64_t size, bool out_int32, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_indices_from_csr_to_coo_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_indices_from_csr_to_coo_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..7035b14199171e197f1fc9d3831d7c320f83c3be
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_indices_from_csr_to_coo_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _convert_indices_from_csr_to_coo(const at::Tensor & crow_indices, const at::Tensor & col_indices, bool out_int32=false, bool transpose=false);
+TORCH_API at::Tensor & _convert_indices_from_csr_to_coo_out(at::Tensor & out, const at::Tensor & crow_indices, const at::Tensor & col_indices, bool out_int32=false, bool transpose=false);
+TORCH_API at::Tensor & _convert_indices_from_csr_to_coo_outf(const at::Tensor & crow_indices, const at::Tensor & col_indices, bool out_int32, bool transpose, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_indices_from_csr_to_coo_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_indices_from_csr_to_coo_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..ca32d3cab9f0ab688edbf8d5e90f33d70ed2104c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_indices_from_csr_to_coo_native.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/_convert_indices_from_csr_to_coo_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured__convert_indices_from_csr_to_coo_structured_cpu : public at::meta::structured__convert_indices_from_csr_to_coo {
+void impl(const at::Tensor & crow_indices, const at::Tensor & col_indices, bool out_int32, bool transpose, const at::Tensor & out);
+};
+struct TORCH_API structured__convert_indices_from_csr_to_coo_structured_cuda : public at::meta::structured__convert_indices_from_csr_to_coo {
+void impl(const at::Tensor & crow_indices, const at::Tensor & col_indices, bool out_int32, bool transpose, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_weight_to_int4pack_for_cpu_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_weight_to_int4pack_for_cpu_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..bc0d88a3293af46c543378bdbc654669f5ceee29
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_weight_to_int4pack_for_cpu_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _convert_weight_to_int4pack_for_cpu(const at::Tensor & self, int64_t innerKTiles);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_weight_to_int4pack_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_weight_to_int4pack_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8189c8eb6359a95dc75e45f2e1d00da969b17637
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convert_weight_to_int4pack_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _convert_weight_to_int4pack {
+  using schema = at::Tensor (const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_convert_weight_to_int4pack";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_convert_weight_to_int4pack(Tensor self, int innerKTiles) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t innerKTiles);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t innerKTiles);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convolution_double_backward_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convolution_double_backward_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..0a38f4f82de5d51ac8dbd5e2049ce5287899e1fd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_convolution_double_backward_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _convolution_double_backward(const ::std::optional<at::Tensor> & ggI, const ::std::optional<at::Tensor> & ggW, const ::std::optional<at::Tensor> & ggb, const at::Tensor & gO, const at::Tensor & weight, const at::Tensor & self, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool transposed, at::IntArrayRef output_padding, int64_t groups, ::std::array<bool,3> output_mask);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _convolution_double_backward_symint(const ::std::optional<at::Tensor> & ggI, const ::std::optional<at::Tensor> & ggW, const ::std::optional<at::Tensor> & ggb, const at::Tensor & gO, const at::Tensor & weight, const at::Tensor & self, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_copy_from_and_resize_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_copy_from_and_resize_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..f6d980b374b4da8e5465ccb3a9b0a7157b65d7f8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_copy_from_and_resize_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _copy_from_and_resize_out(const at::Tensor & self, const at::Tensor & dst, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_ctc_loss_backward.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_ctc_loss_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..9cad37c224e253e8075474f5f0b14053b0744c88
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_ctc_loss_backward.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_ctc_loss_backward_ops.h>
+
+namespace at {
+
+
+// aten::_ctc_loss_backward(Tensor grad, Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, Tensor neg_log_likelihood, Tensor log_alpha, int blank, bool zero_infinity=False) -> Tensor
+inline at::Tensor _ctc_loss_backward(const at::Tensor & grad, const at::Tensor & log_probs, const at::Tensor & targets, at::IntArrayRef input_lengths, at::IntArrayRef target_lengths, const at::Tensor & neg_log_likelihood, const at::Tensor & log_alpha, int64_t blank, bool zero_infinity=false) {
+    return at::_ops::_ctc_loss_backward::call(grad, log_probs, targets, input_lengths, target_lengths, neg_log_likelihood, log_alpha, blank, zero_infinity);
+}
+
+// aten::_ctc_loss_backward.Tensor(Tensor grad, Tensor log_probs, Tensor targets, Tensor input_lengths, Tensor target_lengths, Tensor neg_log_likelihood, Tensor log_alpha, int blank, bool zero_infinity=False) -> Tensor
+inline at::Tensor _ctc_loss_backward(const at::Tensor & grad, const at::Tensor & log_probs, const at::Tensor & targets, const at::Tensor & input_lengths, const at::Tensor & target_lengths, const at::Tensor & neg_log_likelihood, const at::Tensor & log_alpha, int64_t blank, bool zero_infinity=false) {
+    return at::_ops::_ctc_loss_backward_Tensor::call(grad, log_probs, targets, input_lengths, target_lengths, neg_log_likelihood, log_alpha, blank, zero_infinity);
+}
+
+// aten::_ctc_loss_backward.out(Tensor grad, Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, Tensor neg_log_likelihood, Tensor log_alpha, int blank, bool zero_infinity=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _ctc_loss_backward_out(at::Tensor & out, const at::Tensor & grad, const at::Tensor & log_probs, const at::Tensor & targets, at::IntArrayRef input_lengths, at::IntArrayRef target_lengths, const at::Tensor & neg_log_likelihood, const at::Tensor & log_alpha, int64_t blank, bool zero_infinity=false) {
+    return at::_ops::_ctc_loss_backward_out::call(grad, log_probs, targets, input_lengths, target_lengths, neg_log_likelihood, log_alpha, blank, zero_infinity, out);
+}
+// aten::_ctc_loss_backward.out(Tensor grad, Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, Tensor neg_log_likelihood, Tensor log_alpha, int blank, bool zero_infinity=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _ctc_loss_backward_outf(const at::Tensor & grad, const at::Tensor & log_probs, const at::Tensor & targets, at::IntArrayRef input_lengths, at::IntArrayRef target_lengths, const at::Tensor & neg_log_likelihood, const at::Tensor & log_alpha, int64_t blank, bool zero_infinity, at::Tensor & out) {
+    return at::_ops::_ctc_loss_backward_out::call(grad, log_probs, targets, input_lengths, target_lengths, neg_log_likelihood, log_alpha, blank, zero_infinity, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_ctc_loss_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_ctc_loss_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3ab5f11ba794852d76669a09d59e607455cc3cbe
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_ctc_loss_cpu_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> _ctc_loss(const at::Tensor & log_probs, const at::Tensor & targets, at::IntArrayRef input_lengths, at::IntArrayRef target_lengths, int64_t blank=0, bool zero_infinity=false);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> _ctc_loss(const at::Tensor & log_probs, const at::Tensor & targets, const at::Tensor & input_lengths, const at::Tensor & target_lengths, int64_t blank=0, bool zero_infinity=false);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_ctc_loss_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_ctc_loss_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8760eac21dc6e051260a2aee8ad0cc79bdb713ff
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_ctc_loss_meta_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> _ctc_loss(const at::Tensor & log_probs, const at::Tensor & targets, at::IntArrayRef input_lengths, at::IntArrayRef target_lengths, int64_t blank=0, bool zero_infinity=false);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cudnn_ctc_loss_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cudnn_ctc_loss_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..bc60cd4fc6be380297408805233b6bd5b7b640b7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cudnn_ctc_loss_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 &> _cudnn_ctc_loss_out(at::Tensor & out0, at::Tensor & out1, const at::Tensor & log_probs, const at::Tensor & targets, at::IntArrayRef input_lengths, at::IntArrayRef target_lengths, int64_t blank, bool deterministic, bool zero_infinity);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> _cudnn_ctc_loss_outf(const at::Tensor & log_probs, const at::Tensor & targets, at::IntArrayRef input_lengths, at::IntArrayRef target_lengths, int64_t blank, bool deterministic, bool zero_infinity, at::Tensor & out0, at::Tensor & out1);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cufft_clear_plan_cache_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cufft_clear_plan_cache_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e7c95905caa3e9c1c2131306d1ac9223dbaff93e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cufft_clear_plan_cache_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void _cufft_clear_plan_cache(at::DeviceIndex device_index);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cufft_clear_plan_cache_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cufft_clear_plan_cache_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..355a8692d7982c85776e9642796f23a8a9b9c30d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cufft_clear_plan_cache_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _cufft_clear_plan_cache {
+  using schema = void (at::DeviceIndex);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_cufft_clear_plan_cache";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_cufft_clear_plan_cache(DeviceIndex device_index) -> ()";
+  static void call(at::DeviceIndex device_index);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::DeviceIndex device_index);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cummax_helper.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cummax_helper.h
new file mode 100644
index 0000000000000000000000000000000000000000..9508d14d0a9ded2f6440a936b5b95f8f4236ded3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cummax_helper.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_cummax_helper_ops.h>
+
+namespace at {
+
+
+// aten::_cummax_helper(Tensor self, Tensor(a!) values, Tensor(b!) indices, int dim) -> ()
+inline void _cummax_helper(const at::Tensor & self, at::Tensor & values, at::Tensor & indices, int64_t dim) {
+    return at::_ops::_cummax_helper::call(self, values, indices, dim);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cummax_helper_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cummax_helper_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..fa685ddabfedfd6f7a391d0e7da066c5565cd60c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cummax_helper_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API void _cummax_helper(const at::Tensor & self, at::Tensor & values, at::Tensor & indices, int64_t dim);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cummax_helper_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cummax_helper_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..202ca9740cceaae69b64695043633ba920cbfeb3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_cummax_helper_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void _cummax_helper(const at::Tensor & self, at::Tensor & values, at::Tensor & indices, int64_t dim);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_debug_has_internal_overlap.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_debug_has_internal_overlap.h
new file mode 100644
index 0000000000000000000000000000000000000000..73f8cacf914ae011dd2fadac3dcfc5b53964b749
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_debug_has_internal_overlap.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_debug_has_internal_overlap_ops.h>
+
+namespace at {
+
+
+// aten::_debug_has_internal_overlap(Tensor self) -> int
+inline int64_t _debug_has_internal_overlap(const at::Tensor & self) {
+    return at::_ops::_debug_has_internal_overlap::call(self);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_debug_has_internal_overlap_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_debug_has_internal_overlap_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..76dea456fef7f63585533adaf483ad20dbd50aac
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_debug_has_internal_overlap_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 int64_t _debug_has_internal_overlap(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_dyn_quant_pack_4bit_weight_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_dyn_quant_pack_4bit_weight_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..b66d8cf450c726c549cd65e1ed8e193338995e47
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_dyn_quant_pack_4bit_weight_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _dyn_quant_pack_4bit_weight {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, int64_t, int64_t, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_dyn_quant_pack_4bit_weight";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_dyn_quant_pack_4bit_weight(Tensor weights, Tensor scales_zeros, Tensor? bias, int block_size, int in_features, int out_features) -> Tensor";
+  static at::Tensor call(const at::Tensor & weights, const at::Tensor & scales_zeros, const ::std::optional<at::Tensor> & bias, int64_t block_size, int64_t in_features, int64_t out_features);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & weights, const at::Tensor & scales_zeros, const ::std::optional<at::Tensor> & bias, int64_t block_size, int64_t in_features, int64_t out_features);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..3403960e54211b3a33ab374b71ea1f948c8d865e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_backward_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _embedding_bag_backward {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, c10::SymInt, bool, int64_t, bool, const ::std::optional<at::Tensor> &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_embedding_bag_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_embedding_bag_backward(Tensor grad, Tensor indices, Tensor offsets, Tensor offset2bag, Tensor bag_size, Tensor maximum_indices, SymInt num_weights, bool scale_grad_by_freq, int mode, bool sparse, Tensor? per_sample_weights, int padding_idx=-1) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad, const at::Tensor & indices, const at::Tensor & offsets, const at::Tensor & offset2bag, const at::Tensor & bag_size, const at::Tensor & maximum_indices, c10::SymInt num_weights, bool scale_grad_by_freq, int64_t mode, bool sparse, const ::std::optional<at::Tensor> & per_sample_weights, int64_t padding_idx);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad, const at::Tensor & indices, const at::Tensor & offsets, const at::Tensor & offset2bag, const at::Tensor & bag_size, const at::Tensor & maximum_indices, c10::SymInt num_weights, bool scale_grad_by_freq, int64_t mode, bool sparse, const ::std::optional<at::Tensor> & per_sample_weights, int64_t padding_idx);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_dense_backward_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_dense_backward_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..080dfad3fa465a9219a097b99b448859f611f13f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_dense_backward_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _embedding_bag_dense_backward_out(at::Tensor & out, const at::Tensor & grad, const at::Tensor & indices, const at::Tensor & offset2bag, const at::Tensor & bag_size, const at::Tensor & maximum_indices, int64_t num_weights, bool scale_grad_by_freq, int64_t mode, const ::std::optional<at::Tensor> & per_sample_weights, int64_t padding_idx=-1);
+TORCH_API at::Tensor & _embedding_bag_dense_backward_outf(const at::Tensor & grad, const at::Tensor & indices, const at::Tensor & offset2bag, const at::Tensor & bag_size, const at::Tensor & maximum_indices, int64_t num_weights, bool scale_grad_by_freq, int64_t mode, const ::std::optional<at::Tensor> & per_sample_weights, int64_t padding_idx, at::Tensor & out);
+TORCH_API at::Tensor & _embedding_bag_dense_backward_symint_out(at::Tensor & out, const at::Tensor & grad, const at::Tensor & indices, const at::Tensor & offset2bag, const at::Tensor & bag_size, const at::Tensor & maximum_indices, c10::SymInt num_weights, bool scale_grad_by_freq, int64_t mode, const ::std::optional<at::Tensor> & per_sample_weights, int64_t padding_idx=-1);
+TORCH_API at::Tensor & _embedding_bag_dense_backward_symint_outf(const at::Tensor & grad, const at::Tensor & indices, const at::Tensor & offset2bag, const at::Tensor & bag_size, const at::Tensor & maximum_indices, c10::SymInt num_weights, bool scale_grad_by_freq, int64_t mode, const ::std::optional<at::Tensor> & per_sample_weights, int64_t padding_idx, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_forward_only_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_forward_only_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..5bc3634b973850a7ba338f9efb272f15e1acdd27
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_forward_only_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _embedding_bag_forward_only {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, bool, int64_t, bool, const ::std::optional<at::Tensor> &, bool, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_embedding_bag_forward_only";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_embedding_bag_forward_only(Tensor weight, Tensor indices, Tensor offsets, bool scale_grad_by_freq=False, int mode=0, bool sparse=False, Tensor? per_sample_weights=None, bool include_last_offset=False, int padding_idx=-1) -> (Tensor, Tensor, Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> call(const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, bool scale_grad_by_freq, int64_t mode, bool sparse, const ::std::optional<at::Tensor> & per_sample_weights, bool include_last_offset, int64_t padding_idx);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, bool scale_grad_by_freq, int64_t mode, bool sparse, const ::std::optional<at::Tensor> & per_sample_weights, bool include_last_offset, int64_t padding_idx);
+};
+
+struct TORCH_API _embedding_bag_forward_only_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> (const at::Tensor &, const at::Tensor &, const at::Tensor &, bool, int64_t, bool, const ::std::optional<at::Tensor> &, bool, int64_t, at::Tensor &, at::Tensor &, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_embedding_bag_forward_only";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_embedding_bag_forward_only.out(Tensor weight, Tensor indices, Tensor offsets, bool scale_grad_by_freq=False, int mode=0, bool sparse=False, Tensor? per_sample_weights=None, bool include_last_offset=False, int padding_idx=-1, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2, Tensor(d!) out3) -> (Tensor(a!), Tensor(b!), Tensor(c!), Tensor(d!))";
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> call(const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, bool scale_grad_by_freq, int64_t mode, bool sparse, const ::std::optional<at::Tensor> & per_sample_weights, bool include_last_offset, int64_t padding_idx, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3);
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, bool scale_grad_by_freq, int64_t mode, bool sparse, const ::std::optional<at::Tensor> & per_sample_weights, bool include_last_offset, int64_t padding_idx, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..3632a09dae71a07f71cdfb4ff4f072e4ccba08a6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _embedding_bag {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, bool, int64_t, bool, const ::std::optional<at::Tensor> &, bool, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_embedding_bag";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_embedding_bag(Tensor weight, Tensor indices, Tensor offsets, bool scale_grad_by_freq=False, int mode=0, bool sparse=False, Tensor? per_sample_weights=None, bool include_last_offset=False, int padding_idx=-1) -> (Tensor, Tensor, Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> call(const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, bool scale_grad_by_freq, int64_t mode, bool sparse, const ::std::optional<at::Tensor> & per_sample_weights, bool include_last_offset, int64_t padding_idx);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, bool scale_grad_by_freq, int64_t mode, bool sparse, const ::std::optional<at::Tensor> & per_sample_weights, bool include_last_offset, int64_t padding_idx);
+};
+
+struct TORCH_API _embedding_bag_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> (const at::Tensor &, const at::Tensor &, const at::Tensor &, bool, int64_t, bool, const ::std::optional<at::Tensor> &, bool, int64_t, at::Tensor &, at::Tensor &, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_embedding_bag";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_embedding_bag.out(Tensor weight, Tensor indices, Tensor offsets, bool scale_grad_by_freq=False, int mode=0, bool sparse=False, Tensor? per_sample_weights=None, bool include_last_offset=False, int padding_idx=-1, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2, Tensor(d!) out3) -> (Tensor(a!), Tensor(b!), Tensor(c!), Tensor(d!))";
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> call(const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, bool scale_grad_by_freq, int64_t mode, bool sparse, const ::std::optional<at::Tensor> & per_sample_weights, bool include_last_offset, int64_t padding_idx, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3);
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, bool scale_grad_by_freq, int64_t mode, bool sparse, const ::std::optional<at::Tensor> & per_sample_weights, bool include_last_offset, int64_t padding_idx, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_sparse_backward.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_sparse_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..89e04000d8ec2609bcb37cc9c9bb7d4a9ba70123
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_embedding_bag_sparse_backward.h
@@ -0,0 +1,53 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_embedding_bag_sparse_backward_ops.h>
+
+namespace at {
+
+
+// aten::_embedding_bag_sparse_backward(Tensor grad, Tensor indices, Tensor offsets, Tensor offset2bag, Tensor bag_size, SymInt num_weights, bool scale_grad_by_freq, int mode, Tensor? per_sample_weights, int padding_idx=-1) -> Tensor
+inline at::Tensor _embedding_bag_sparse_backward(const at::Tensor & grad, const at::Tensor & indices, const at::Tensor & offsets, const at::Tensor & offset2bag, const at::Tensor & bag_size, int64_t num_weights, bool scale_grad_by_freq, int64_t mode, const ::std::optional<at::Tensor> & per_sample_weights, int64_t padding_idx=-1) {
+    return at::_ops::_embedding_bag_sparse_backward::call(grad, indices, offsets, offset2bag, bag_size, num_weights, scale_grad_by_freq, mode, per_sample_weights, padding_idx);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor _embedding_bag_sparse_backward(const at::Tensor & grad, const at::Tensor & indices, const at::Tensor & offsets, const at::Tensor & offset2bag, const at::Tensor & bag_size, int64_t num_weights, bool scale_grad_by_freq, int64_t mode, const ::std::optional<at::Tensor> & per_sample_weights, int64_t padding_idx=-1) {
+    return at::_ops::_embedding_bag_sparse_backward::call(grad, indices, offsets, offset2bag, bag_size, num_weights, scale_grad_by_freq, mode, per_sample_weights, padding_idx);
+  }
+}
+
+// aten::_embedding_bag_sparse_backward(Tensor grad, Tensor indices, Tensor offsets, Tensor offset2bag, Tensor bag_size, SymInt num_weights, bool scale_grad_by_freq, int mode, Tensor? per_sample_weights, int padding_idx=-1) -> Tensor
+inline at::Tensor _embedding_bag_sparse_backward_symint(const at::Tensor & grad, const at::Tensor & indices, const at::Tensor & offsets, const at::Tensor & offset2bag, const at::Tensor & bag_size, c10::SymInt num_weights, bool scale_grad_by_freq, int64_t mode, const ::std::optional<at::Tensor> & per_sample_weights, int64_t padding_idx=-1) {
+    return at::_ops::_embedding_bag_sparse_backward::call(grad, indices, offsets, offset2bag, bag_size, num_weights, scale_grad_by_freq, mode, per_sample_weights, padding_idx);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor _embedding_bag_sparse_backward(const at::Tensor & grad, const at::Tensor & indices, const at::Tensor & offsets, const at::Tensor & offset2bag, const at::Tensor & bag_size, c10::SymInt num_weights, bool scale_grad_by_freq, int64_t mode, const ::std::optional<at::Tensor> & per_sample_weights, int64_t padding_idx=-1) {
+    return at::_ops::_embedding_bag_sparse_backward::call(grad, indices, offsets, offset2bag, bag_size, num_weights, scale_grad_by_freq, mode, per_sample_weights, padding_idx);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_empty_affine_quantized.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_empty_affine_quantized.h
new file mode 100644
index 0000000000000000000000000000000000000000..e057b95ce6860292bb51526bcbb0d3bc2aad5016
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_empty_affine_quantized.h
@@ -0,0 +1,119 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_empty_affine_quantized_ops.h>
+
+namespace at {
+
+
+// aten::_empty_affine_quantized(SymInt[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, float scale=1, int zero_point=0, MemoryFormat? memory_format=contiguous_format) -> Tensor
+inline at::Tensor _empty_affine_quantized(at::IntArrayRef size, at::TensorOptions options={}, double scale=1, int64_t zero_point=0, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous) {
+    return at::_ops::_empty_affine_quantized::call(c10::fromIntArrayRefSlow(size), c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt(), scale, zero_point, c10::impl::check_tensor_options_and_extract_memory_format(options, memory_format));
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor _empty_affine_quantized(at::IntArrayRef size, at::TensorOptions options={}, double scale=1, int64_t zero_point=0, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous) {
+    return at::_ops::_empty_affine_quantized::call(c10::fromIntArrayRefSlow(size), c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt(), scale, zero_point, c10::impl::check_tensor_options_and_extract_memory_format(options, memory_format));
+  }
+}
+
+// aten::_empty_affine_quantized(SymInt[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, float scale=1, int zero_point=0, MemoryFormat? memory_format=contiguous_format) -> Tensor
+inline at::Tensor _empty_affine_quantized(at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, double scale, int64_t zero_point, ::std::optional<at::MemoryFormat> memory_format) {
+    return at::_ops::_empty_affine_quantized::call(c10::fromIntArrayRefSlow(size), dtype, layout, device, pin_memory, scale, zero_point, memory_format);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor _empty_affine_quantized(at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, double scale, int64_t zero_point, ::std::optional<at::MemoryFormat> memory_format) {
+    return at::_ops::_empty_affine_quantized::call(c10::fromIntArrayRefSlow(size), dtype, layout, device, pin_memory, scale, zero_point, memory_format);
+  }
+}
+
+// aten::_empty_affine_quantized(SymInt[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, float scale=1, int zero_point=0, MemoryFormat? memory_format=contiguous_format) -> Tensor
+inline at::Tensor _empty_affine_quantized_symint(c10::SymIntArrayRef size, at::TensorOptions options={}, double scale=1, int64_t zero_point=0, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous) {
+    return at::_ops::_empty_affine_quantized::call(size, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt(), scale, zero_point, c10::impl::check_tensor_options_and_extract_memory_format(options, memory_format));
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor _empty_affine_quantized(c10::SymIntArrayRef size, at::TensorOptions options={}, double scale=1, int64_t zero_point=0, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous) {
+    return at::_ops::_empty_affine_quantized::call(size, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt(), scale, zero_point, c10::impl::check_tensor_options_and_extract_memory_format(options, memory_format));
+  }
+}
+
+// aten::_empty_affine_quantized(SymInt[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, float scale=1, int zero_point=0, MemoryFormat? memory_format=contiguous_format) -> Tensor
+inline at::Tensor _empty_affine_quantized_symint(c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, double scale, int64_t zero_point, ::std::optional<at::MemoryFormat> memory_format) {
+    return at::_ops::_empty_affine_quantized::call(size, dtype, layout, device, pin_memory, scale, zero_point, memory_format);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor _empty_affine_quantized(c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, double scale, int64_t zero_point, ::std::optional<at::MemoryFormat> memory_format) {
+    return at::_ops::_empty_affine_quantized::call(size, dtype, layout, device, pin_memory, scale, zero_point, memory_format);
+  }
+}
+
+// aten::_empty_affine_quantized.out(SymInt[] size, *, float scale=1, int zero_point=0, MemoryFormat? memory_format=contiguous_format, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _empty_affine_quantized_out(at::Tensor & out, at::IntArrayRef size, double scale=1, int64_t zero_point=0, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous) {
+    return at::_ops::_empty_affine_quantized_out::call(c10::fromIntArrayRefSlow(size), scale, zero_point, memory_format, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & _empty_affine_quantized_out(at::Tensor & out, at::IntArrayRef size, double scale=1, int64_t zero_point=0, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous) {
+    return at::_ops::_empty_affine_quantized_out::call(c10::fromIntArrayRefSlow(size), scale, zero_point, memory_format, out);
+  }
+}
+
+// aten::_empty_affine_quantized.out(SymInt[] size, *, float scale=1, int zero_point=0, MemoryFormat? memory_format=contiguous_format, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _empty_affine_quantized_outf(at::IntArrayRef size, double scale, int64_t zero_point, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out) {
+    return at::_ops::_empty_affine_quantized_out::call(c10::fromIntArrayRefSlow(size), scale, zero_point, memory_format, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & _empty_affine_quantized_outf(at::IntArrayRef size, double scale, int64_t zero_point, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out) {
+    return at::_ops::_empty_affine_quantized_out::call(c10::fromIntArrayRefSlow(size), scale, zero_point, memory_format, out);
+  }
+}
+
+// aten::_empty_affine_quantized.out(SymInt[] size, *, float scale=1, int zero_point=0, MemoryFormat? memory_format=contiguous_format, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _empty_affine_quantized_symint_out(at::Tensor & out, c10::SymIntArrayRef size, double scale=1, int64_t zero_point=0, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous) {
+    return at::_ops::_empty_affine_quantized_out::call(size, scale, zero_point, memory_format, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & _empty_affine_quantized_out(at::Tensor & out, c10::SymIntArrayRef size, double scale=1, int64_t zero_point=0, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous) {
+    return at::_ops::_empty_affine_quantized_out::call(size, scale, zero_point, memory_format, out);
+  }
+}
+
+// aten::_empty_affine_quantized.out(SymInt[] size, *, float scale=1, int zero_point=0, MemoryFormat? memory_format=contiguous_format, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _empty_affine_quantized_symint_outf(c10::SymIntArrayRef size, double scale, int64_t zero_point, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out) {
+    return at::_ops::_empty_affine_quantized_out::call(size, scale, zero_point, memory_format, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & _empty_affine_quantized_outf(c10::SymIntArrayRef size, double scale, int64_t zero_point, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out) {
+    return at::_ops::_empty_affine_quantized_out::call(size, scale, zero_point, memory_format, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_empty_affine_quantized_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_empty_affine_quantized_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..fdac4039218f036227f30964e210afeb524359f6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_empty_affine_quantized_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _empty_affine_quantized_out_symint(c10::SymIntArrayRef size, double scale, int64_t zero_point, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor empty_affine_quantized_other_backends_stub(at::IntArrayRef size, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={}, double scale=1, int64_t zero_point=0, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous);
+TORCH_API at::Tensor empty_affine_quantized(at::IntArrayRef size, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={}, double scale=1, int64_t zero_point=0, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_empty_per_channel_affine_quantized_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_empty_per_channel_affine_quantized_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..f2a416942e6b8fd9997faa8ee21b39d776e24345
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_empty_per_channel_affine_quantized_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _empty_per_channel_affine_quantized_out_symint(c10::SymIntArrayRef 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_other_backends_stub(at::IntArrayRef size, const at::Tensor & scales, const at::Tensor & zero_points, int64_t axis, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={}, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous);
+TORCH_API at::Tensor empty_per_channel_affine_quantized(at::IntArrayRef size, const at::Tensor & scales, const at::Tensor & zero_points, int64_t axis, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={}, ::std::optional<at::MemoryFormat> memory_format=c10::MemoryFormat::Contiguous);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_euclidean_dist_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_euclidean_dist_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ddf765d22f899df3a0adbfa3f92254a3997897f0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_euclidean_dist_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor _euclidean_dist(const at::Tensor & x1, const at::Tensor & x2);
+TORCH_API at::Tensor & _euclidean_dist_out(at::Tensor & out, const at::Tensor & x1, const at::Tensor & x2);
+TORCH_API at::Tensor & _euclidean_dist_outf(const at::Tensor & x1, const at::Tensor & x2, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fake_quantize_per_tensor_affine_cachemask_tensor_qparams_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fake_quantize_per_tensor_affine_cachemask_tensor_qparams_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..2824af81e46863d4d3529011f5ed20ab9bc8a603
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fake_quantize_per_tensor_affine_cachemask_tensor_qparams_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> _fake_quantize_per_tensor_affine_cachemask_tensor_qparams(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, const at::Tensor & fake_quant_enabled, int64_t quant_min, int64_t quant_max);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fft_c2c_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fft_c2c_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d2115217f279267a31eec4099b42ef5e40cba30e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fft_c2c_cpu_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _fft_c2c(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool forward);
+TORCH_API at::Tensor _fft_c2c_symint(const at::Tensor & self, c10::SymIntArrayRef dim, int64_t normalization, bool forward);
+TORCH_API at::Tensor & _fft_c2c_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool forward);
+TORCH_API at::Tensor & _fft_c2c_outf(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool forward, at::Tensor & out);
+TORCH_API at::Tensor & _fft_c2c_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef dim, int64_t normalization, bool forward);
+TORCH_API at::Tensor & _fft_c2c_symint_outf(const at::Tensor & self, c10::SymIntArrayRef dim, int64_t normalization, bool forward, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fft_c2c_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fft_c2c_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..866253752563c7ae2955a73e71d7f9e0ae3995c1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fft_c2c_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_c2c_mkl(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool forward);
+TORCH_API at::Tensor & _fft_c2c_mkl_out(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool forward, at::Tensor & out);
+TORCH_API at::Tensor _fft_c2c_cufft(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool forward);
+TORCH_API at::Tensor & _fft_c2c_cufft_out(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool forward, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fft_c2c_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fft_c2c_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..0a1e3c720c5c3956b06aaa4d02c81cb3dcb2b45c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fft_c2c_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _fft_c2c {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef, int64_t, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_fft_c2c";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_fft_c2c(Tensor self, SymInt[] dim, int normalization, bool forward) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymIntArrayRef dim, int64_t normalization, bool forward);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef dim, int64_t normalization, bool forward);
+};
+
+struct TORCH_API _fft_c2c_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymIntArrayRef, int64_t, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_fft_c2c";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_fft_c2c.out(Tensor self, SymInt[] dim, int normalization, bool forward, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymIntArrayRef dim, int64_t normalization, bool forward, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef dim, int64_t normalization, bool forward, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foobar_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foobar_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3a9d88faa9b705beec4a478c29c836c8e5ba48d3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foobar_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _foobar(const at::Tensor & self, bool arg1=true, bool arg2=true, bool arg3=true);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_acos_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_acos_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..5c2113c2498532be1181b457365d02136033156d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_acos_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_acos {
+  using schema = ::std::vector<at::Tensor> (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_acos";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_acos(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_acos_ {
+  using schema = void (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_acos_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_acos_(Tensor(a!)[] self) -> ()";
+  static void call(at::TensorList self);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_acos_out {
+  using schema = void (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_acos";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_foreach_acos.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_asin_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_asin_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..bfca355ddd7c7ac8f9dcfa3995e4a72e9254cd42
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_asin_native.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_asin_slow(at::TensorList self);
+TORCH_API void _foreach_asin_out(at::TensorList self, at::TensorList out);
+TORCH_API void foreach_tensor_asin_slow_(at::TensorList self);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_asin_cuda(at::TensorList self);
+TORCH_API void foreach_tensor_asin_cuda_(at::TensorList self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_asin_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_asin_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..0bf722dbaa99f78ca5cae705cbf210d2f0fcdf2f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_asin_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _foreach_asin {
+  using schema = ::std::vector<at::Tensor> (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_asin";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_asin(Tensor[] self) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList self);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_asin_ {
+  using schema = void (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_asin_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_asin_(Tensor(a!)[] self) -> ()";
+  static void call(at::TensorList self);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_asin_out {
+  using schema = void (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_asin";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_foreach_asin.out(Tensor[] self, *, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList self, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_copy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_copy.h
new file mode 100644
index 0000000000000000000000000000000000000000..eca5d87e4c757f946fc446cf4bb6df313f30b5c1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_copy.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_foreach_copy_ops.h>
+
+namespace at {
+
+
+// aten::_foreach_copy_(Tensor(a!)[] self, Tensor[] src, bool non_blocking=False) -> ()
+inline void _foreach_copy_(at::TensorList self, at::TensorList src, bool non_blocking=false) {
+    return at::_ops::_foreach_copy_::call(self, src, non_blocking);
+}
+
+// aten::_foreach_copy(Tensor[] self, Tensor[] src, bool non_blocking=False) -> Tensor[] self_out
+inline ::std::vector<at::Tensor> _foreach_copy(at::TensorList self, at::TensorList src, bool non_blocking=false) {
+    return at::_ops::_foreach_copy::call(self, src, non_blocking);
+}
+
+// aten::_foreach_copy.out(Tensor[] self, Tensor[] src, bool non_blocking=False, *, Tensor(a!)[] out) -> ()
+inline void _foreach_copy_out(at::TensorList out, at::TensorList self, at::TensorList src, bool non_blocking=false) {
+    return at::_ops::_foreach_copy_out::call(self, src, non_blocking, out);
+}
+// aten::_foreach_copy.out(Tensor[] self, Tensor[] src, bool non_blocking=False, *, Tensor(a!)[] out) -> ()
+inline void _foreach_copy_outf(at::TensorList self, at::TensorList src, bool non_blocking, at::TensorList out) {
+    return at::_ops::_foreach_copy_out::call(self, src, non_blocking, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_cos_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_cos_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..44782687ccd7c4467977cab344c06dbadb2eafee
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_cos_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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::vector<at::Tensor> _foreach_cos(at::TensorList self);
+TORCH_API void _foreach_cos_out(at::TensorList out, at::TensorList self);
+TORCH_API void _foreach_cos_outf(at::TensorList self, at::TensorList out);
+TORCH_API void _foreach_cos_(at::TensorList self);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_cos_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_cos_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..05f3b102bb43fa26ab2bee8cb8331446faac491e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_cos_cuda_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API ::std::vector<at::Tensor> _foreach_cos(at::TensorList self);
+TORCH_API void _foreach_cos_(at::TensorList self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_div_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_div_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..376daa0475e7949da9e6b052c28b25b63dbdf6af
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_div_ops.h
@@ -0,0 +1,155 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_div_Scalar {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "_foreach_div.Scalar(Tensor[] self, Scalar scalar) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList self, const at::Scalar & scalar);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, const at::Scalar & scalar);
+};
+
+struct TORCH_API _foreach_div__Scalar {
+  using schema = void (at::TensorList, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div_";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "_foreach_div_.Scalar(Tensor(a!)[] self, Scalar scalar) -> ()";
+  static void call(at::TensorList self, const at::Scalar & scalar);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, const at::Scalar & scalar);
+};
+
+struct TORCH_API _foreach_div_List {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div";
+  static constexpr const char* overload_name = "List";
+  static constexpr const char* schema_str = "_foreach_div.List(Tensor[] self, Tensor[] other) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList self, at::TensorList other);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList other);
+};
+
+struct TORCH_API _foreach_div__List {
+  using schema = void (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div_";
+  static constexpr const char* overload_name = "List";
+  static constexpr const char* schema_str = "_foreach_div_.List(Tensor(a!)[] self, Tensor[] other) -> ()";
+  static void call(at::TensorList self, at::TensorList other);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList other);
+};
+
+struct TORCH_API _foreach_div_ScalarList {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, at::ArrayRef<at::Scalar>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div";
+  static constexpr const char* overload_name = "ScalarList";
+  static constexpr const char* schema_str = "_foreach_div.ScalarList(Tensor[] self, Scalar[] scalars) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+};
+
+struct TORCH_API _foreach_div__ScalarList {
+  using schema = void (at::TensorList, at::ArrayRef<at::Scalar>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div_";
+  static constexpr const char* overload_name = "ScalarList";
+  static constexpr const char* schema_str = "_foreach_div_.ScalarList(Tensor(a!)[] self, Scalar[] scalars) -> ()";
+  static void call(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+};
+
+struct TORCH_API _foreach_div_Tensor {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "_foreach_div.Tensor(Tensor[] self, Tensor other) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList self, const at::Tensor & other);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, const at::Tensor & other);
+};
+
+struct TORCH_API _foreach_div__Tensor {
+  using schema = void (at::TensorList, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div_";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "_foreach_div_.Tensor(Tensor(a!)[] self, Tensor other) -> ()";
+  static void call(at::TensorList self, const at::Tensor & other);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, const at::Tensor & other);
+};
+
+struct TORCH_API _foreach_div_Scalar_out {
+  using schema = void (at::TensorList, const at::Scalar &, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div";
+  static constexpr const char* overload_name = "Scalar_out";
+  static constexpr const char* schema_str = "_foreach_div.Scalar_out(Tensor[] self, Scalar scalar, *, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList self, const at::Scalar & scalar, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, const at::Scalar & scalar, at::TensorList out);
+};
+
+struct TORCH_API _foreach_div_List_out {
+  using schema = void (at::TensorList, at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div";
+  static constexpr const char* overload_name = "List_out";
+  static constexpr const char* schema_str = "_foreach_div.List_out(Tensor[] self, Tensor[] other, *, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList self, at::TensorList other, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList other, at::TensorList out);
+};
+
+struct TORCH_API _foreach_div_ScalarList_out {
+  using schema = void (at::TensorList, at::ArrayRef<at::Scalar>, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div";
+  static constexpr const char* overload_name = "ScalarList_out";
+  static constexpr const char* schema_str = "_foreach_div.ScalarList_out(Tensor[] self, Scalar[] scalars, *, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList self, at::ArrayRef<at::Scalar> scalars, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::ArrayRef<at::Scalar> scalars, at::TensorList out);
+};
+
+struct TORCH_API _foreach_div_Tensor_out {
+  using schema = void (at::TensorList, const at::Tensor &, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_div";
+  static constexpr const char* overload_name = "Tensor_out";
+  static constexpr const char* schema_str = "_foreach_div.Tensor_out(Tensor[] self, Tensor other, *, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList self, const at::Tensor & other, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, const at::Tensor & other, at::TensorList out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf.h
new file mode 100644
index 0000000000000000000000000000000000000000..ea8ec05c4e2fe592c98deac3f5aa08189da2c5e4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_foreach_erf_ops.h>
+
+namespace at {
+
+
+// aten::_foreach_erf(Tensor[] self) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_erf(at::TensorList self) {
+    return at::_ops::_foreach_erf::call(self);
+}
+
+// aten::_foreach_erf_(Tensor(a!)[] self) -> ()
+inline void _foreach_erf_(at::TensorList self) {
+    return at::_ops::_foreach_erf_::call(self);
+}
+
+// aten::_foreach_erf.out(Tensor[] self, *, Tensor(a!)[] out) -> ()
+inline void _foreach_erf_out(at::TensorList out, at::TensorList self) {
+    return at::_ops::_foreach_erf_out::call(self, out);
+}
+// aten::_foreach_erf.out(Tensor[] self, *, Tensor(a!)[] out) -> ()
+inline void _foreach_erf_outf(at::TensorList self, at::TensorList out) {
+    return at::_ops::_foreach_erf_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..185bc138b31856ab2b1078b063c745da0f74eb80
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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::vector<at::Tensor> _foreach_erf(at::TensorList self);
+TORCH_API void _foreach_erf_out(at::TensorList out, at::TensorList self);
+TORCH_API void _foreach_erf_outf(at::TensorList self, at::TensorList out);
+TORCH_API void _foreach_erf_(at::TensorList self);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..892d6eea91628c3c98043082833b7ee4b36c710c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf_native.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_erf_slow(at::TensorList self);
+TORCH_API void _foreach_erf_out(at::TensorList self, at::TensorList out);
+TORCH_API void foreach_tensor_erf_slow_(at::TensorList self);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_erf_cuda(at::TensorList self);
+TORCH_API void foreach_tensor_erf_cuda_(at::TensorList self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8e9ff74e133db2b4853a2130cfb42dc9a1d43e4a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_erf_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_erf {
+  using schema = ::std::vector<at::Tensor> (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_erf";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_erf(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_erf_ {
+  using schema = void (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_erf_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_erf_(Tensor(a!)[] self) -> ()";
+  static void call(at::TensorList self);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_erf_out {
+  using schema = void (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_erf";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_foreach_erf.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_exp_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_exp_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..22d032c7d5fb8327bf4007ab5b7009d69a5c97e8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_exp_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_exp {
+  using schema = ::std::vector<at::Tensor> (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_exp";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_exp(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_exp_ {
+  using schema = void (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_exp_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_exp_(Tensor(a!)[] self) -> ()";
+  static void call(at::TensorList self);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_exp_out {
+  using schema = void (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_exp";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_foreach_exp.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_lerp.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_lerp.h
new file mode 100644
index 0000000000000000000000000000000000000000..3e3b604a83a59301a08bc967552ab9c799c46285
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_lerp.h
@@ -0,0 +1,88 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_foreach_lerp_ops.h>
+
+namespace at {
+
+
+// aten::_foreach_lerp.List(Tensor[] self, Tensor[] tensors1, Tensor[] weights) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_lerp(at::TensorList self, at::TensorList tensors1, at::TensorList weights) {
+    return at::_ops::_foreach_lerp_List::call(self, tensors1, weights);
+}
+
+// aten::_foreach_lerp_.List(Tensor(a!)[] self, Tensor[] tensors1, Tensor[] weights) -> ()
+inline void _foreach_lerp_(at::TensorList self, at::TensorList tensors1, at::TensorList weights) {
+    return at::_ops::_foreach_lerp__List::call(self, tensors1, weights);
+}
+
+// aten::_foreach_lerp.Scalar(Tensor[] self, Tensor[] tensors1, Scalar weight) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_lerp(at::TensorList self, at::TensorList tensors1, const at::Scalar & weight) {
+    return at::_ops::_foreach_lerp_Scalar::call(self, tensors1, weight);
+}
+
+// aten::_foreach_lerp_.Scalar(Tensor(a!)[] self, Tensor[] tensors1, Scalar weight) -> ()
+inline void _foreach_lerp_(at::TensorList self, at::TensorList tensors1, const at::Scalar & weight) {
+    return at::_ops::_foreach_lerp__Scalar::call(self, tensors1, weight);
+}
+
+// aten::_foreach_lerp.ScalarList(Tensor[] self, Tensor[] tensors1, Scalar[] weight) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_lerp(at::TensorList self, at::TensorList tensors1, at::ArrayRef<at::Scalar> weight) {
+    return at::_ops::_foreach_lerp_ScalarList::call(self, tensors1, weight);
+}
+
+// aten::_foreach_lerp_.ScalarList(Tensor(a!)[] self, Tensor[] tensors1, Scalar[] weight) -> ()
+inline void _foreach_lerp_(at::TensorList self, at::TensorList tensors1, at::ArrayRef<at::Scalar> weight) {
+    return at::_ops::_foreach_lerp__ScalarList::call(self, tensors1, weight);
+}
+
+// aten::_foreach_lerp.List_out(Tensor[] self, Tensor[] tensors1, Tensor[] weights, *, Tensor(a!)[] out) -> ()
+inline void _foreach_lerp_out(at::TensorList out, at::TensorList self, at::TensorList tensors1, at::TensorList weights) {
+    return at::_ops::_foreach_lerp_List_out::call(self, tensors1, weights, out);
+}
+// aten::_foreach_lerp.List_out(Tensor[] self, Tensor[] tensors1, Tensor[] weights, *, Tensor(a!)[] out) -> ()
+inline void _foreach_lerp_outf(at::TensorList self, at::TensorList tensors1, at::TensorList weights, at::TensorList out) {
+    return at::_ops::_foreach_lerp_List_out::call(self, tensors1, weights, out);
+}
+
+// aten::_foreach_lerp.Scalar_out(Tensor[] self, Tensor[] tensors1, Scalar weight, *, Tensor(a!)[] out) -> ()
+inline void _foreach_lerp_out(at::TensorList out, at::TensorList self, at::TensorList tensors1, const at::Scalar & weight) {
+    return at::_ops::_foreach_lerp_Scalar_out::call(self, tensors1, weight, out);
+}
+// aten::_foreach_lerp.Scalar_out(Tensor[] self, Tensor[] tensors1, Scalar weight, *, Tensor(a!)[] out) -> ()
+inline void _foreach_lerp_outf(at::TensorList self, at::TensorList tensors1, const at::Scalar & weight, at::TensorList out) {
+    return at::_ops::_foreach_lerp_Scalar_out::call(self, tensors1, weight, out);
+}
+
+// aten::_foreach_lerp.ScalarList_out(Tensor[] self, Tensor[] tensors1, Scalar[] weight, *, Tensor(a!)[] out) -> ()
+inline void _foreach_lerp_out(at::TensorList out, at::TensorList self, at::TensorList tensors1, at::ArrayRef<at::Scalar> weight) {
+    return at::_ops::_foreach_lerp_ScalarList_out::call(self, tensors1, weight, out);
+}
+// aten::_foreach_lerp.ScalarList_out(Tensor[] self, Tensor[] tensors1, Scalar[] weight, *, Tensor(a!)[] out) -> ()
+inline void _foreach_lerp_outf(at::TensorList self, at::TensorList tensors1, at::ArrayRef<at::Scalar> weight, at::TensorList out) {
+    return at::_ops::_foreach_lerp_ScalarList_out::call(self, tensors1, weight, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_lerp_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_lerp_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..6b5ef0aec636dc312ab444db84db4ba28737abf8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_lerp_ops.h
@@ -0,0 +1,122 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_lerp_List {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_lerp";
+  static constexpr const char* overload_name = "List";
+  static constexpr const char* schema_str = "_foreach_lerp.List(Tensor[] self, Tensor[] tensors1, Tensor[] weights) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList self, at::TensorList tensors1, at::TensorList weights);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensors1, at::TensorList weights);
+};
+
+struct TORCH_API _foreach_lerp__List {
+  using schema = void (at::TensorList, at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_lerp_";
+  static constexpr const char* overload_name = "List";
+  static constexpr const char* schema_str = "_foreach_lerp_.List(Tensor(a!)[] self, Tensor[] tensors1, Tensor[] weights) -> ()";
+  static void call(at::TensorList self, at::TensorList tensors1, at::TensorList weights);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensors1, at::TensorList weights);
+};
+
+struct TORCH_API _foreach_lerp_Scalar {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, at::TensorList, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_lerp";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "_foreach_lerp.Scalar(Tensor[] self, Tensor[] tensors1, Scalar weight) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList self, at::TensorList tensors1, const at::Scalar & weight);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensors1, const at::Scalar & weight);
+};
+
+struct TORCH_API _foreach_lerp__Scalar {
+  using schema = void (at::TensorList, at::TensorList, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_lerp_";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "_foreach_lerp_.Scalar(Tensor(a!)[] self, Tensor[] tensors1, Scalar weight) -> ()";
+  static void call(at::TensorList self, at::TensorList tensors1, const at::Scalar & weight);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensors1, const at::Scalar & weight);
+};
+
+struct TORCH_API _foreach_lerp_ScalarList {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, at::TensorList, at::ArrayRef<at::Scalar>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_lerp";
+  static constexpr const char* overload_name = "ScalarList";
+  static constexpr const char* schema_str = "_foreach_lerp.ScalarList(Tensor[] self, Tensor[] tensors1, Scalar[] weight) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList self, at::TensorList tensors1, at::ArrayRef<at::Scalar> weight);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensors1, at::ArrayRef<at::Scalar> weight);
+};
+
+struct TORCH_API _foreach_lerp__ScalarList {
+  using schema = void (at::TensorList, at::TensorList, at::ArrayRef<at::Scalar>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_lerp_";
+  static constexpr const char* overload_name = "ScalarList";
+  static constexpr const char* schema_str = "_foreach_lerp_.ScalarList(Tensor(a!)[] self, Tensor[] tensors1, Scalar[] weight) -> ()";
+  static void call(at::TensorList self, at::TensorList tensors1, at::ArrayRef<at::Scalar> weight);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensors1, at::ArrayRef<at::Scalar> weight);
+};
+
+struct TORCH_API _foreach_lerp_List_out {
+  using schema = void (at::TensorList, at::TensorList, at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_lerp";
+  static constexpr const char* overload_name = "List_out";
+  static constexpr const char* schema_str = "_foreach_lerp.List_out(Tensor[] self, Tensor[] tensors1, Tensor[] weights, *, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList self, at::TensorList tensors1, at::TensorList weights, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensors1, at::TensorList weights, at::TensorList out);
+};
+
+struct TORCH_API _foreach_lerp_Scalar_out {
+  using schema = void (at::TensorList, at::TensorList, const at::Scalar &, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_lerp";
+  static constexpr const char* overload_name = "Scalar_out";
+  static constexpr const char* schema_str = "_foreach_lerp.Scalar_out(Tensor[] self, Tensor[] tensors1, Scalar weight, *, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList self, at::TensorList tensors1, const at::Scalar & weight, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensors1, const at::Scalar & weight, at::TensorList out);
+};
+
+struct TORCH_API _foreach_lerp_ScalarList_out {
+  using schema = void (at::TensorList, at::TensorList, at::ArrayRef<at::Scalar>, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_lerp";
+  static constexpr const char* overload_name = "ScalarList_out";
+  static constexpr const char* schema_str = "_foreach_lerp.ScalarList_out(Tensor[] self, Tensor[] tensors1, Scalar[] weight, *, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList self, at::TensorList tensors1, at::ArrayRef<at::Scalar> weight, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensors1, at::ArrayRef<at::Scalar> weight, at::TensorList out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_log.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_log.h
new file mode 100644
index 0000000000000000000000000000000000000000..97f5c5fb0da1dbd7dc3ad93a9809faee337b7d29
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_log.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_foreach_log_ops.h>
+
+namespace at {
+
+
+// aten::_foreach_log(Tensor[] self) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_log(at::TensorList self) {
+    return at::_ops::_foreach_log::call(self);
+}
+
+// aten::_foreach_log_(Tensor(a!)[] self) -> ()
+inline void _foreach_log_(at::TensorList self) {
+    return at::_ops::_foreach_log_::call(self);
+}
+
+// aten::_foreach_log.out(Tensor[] self, *, Tensor(a!)[] out) -> ()
+inline void _foreach_log_out(at::TensorList out, at::TensorList self) {
+    return at::_ops::_foreach_log_out::call(self, out);
+}
+// aten::_foreach_log.out(Tensor[] self, *, Tensor(a!)[] out) -> ()
+inline void _foreach_log_outf(at::TensorList self, at::TensorList out) {
+    return at::_ops::_foreach_log_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_log1p_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_log1p_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..79fc0745674bf030b0ca713709d0641e12f804e8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_log1p_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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::vector<at::Tensor> _foreach_log1p(at::TensorList self);
+TORCH_API void _foreach_log1p_out(at::TensorList out, at::TensorList self);
+TORCH_API void _foreach_log1p_outf(at::TensorList self, at::TensorList out);
+TORCH_API void _foreach_log1p_(at::TensorList self);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_log1p_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_log1p_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..d9e789da87cdf027fa8321f229dbeeece7d8364c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_log1p_native.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_log1p_slow(at::TensorList self);
+TORCH_API void _foreach_log1p_out(at::TensorList self, at::TensorList out);
+TORCH_API void foreach_tensor_log1p_slow_(at::TensorList self);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_log1p_cuda(at::TensorList self);
+TORCH_API void foreach_tensor_log1p_cuda_(at::TensorList self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_max_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_max_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..20e1172ec9b111a894f279b6d8cfb77e2b82f1bc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_max_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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::vector<at::Tensor> _foreach_max(at::TensorList self);
+TORCH_API void _foreach_max_out(at::TensorList out, at::TensorList self);
+TORCH_API void _foreach_max_outf(at::TensorList self, at::TensorList out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_maximum.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_maximum.h
new file mode 100644
index 0000000000000000000000000000000000000000..67edb9b2d3523fde26f8b817455d53b47d5e18cd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_maximum.h
@@ -0,0 +1,88 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_foreach_maximum_ops.h>
+
+namespace at {
+
+
+// aten::_foreach_maximum.Scalar(Tensor[] self, Scalar scalar) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_maximum(at::TensorList self, const at::Scalar & scalar) {
+    return at::_ops::_foreach_maximum_Scalar::call(self, scalar);
+}
+
+// aten::_foreach_maximum_.Scalar(Tensor(a!)[] self, Scalar scalar) -> ()
+inline void _foreach_maximum_(at::TensorList self, const at::Scalar & scalar) {
+    return at::_ops::_foreach_maximum__Scalar::call(self, scalar);
+}
+
+// aten::_foreach_maximum.List(Tensor[] self, Tensor[] other) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_maximum(at::TensorList self, at::TensorList other) {
+    return at::_ops::_foreach_maximum_List::call(self, other);
+}
+
+// aten::_foreach_maximum_.List(Tensor(a!)[] self, Tensor[] other) -> ()
+inline void _foreach_maximum_(at::TensorList self, at::TensorList other) {
+    return at::_ops::_foreach_maximum__List::call(self, other);
+}
+
+// aten::_foreach_maximum.ScalarList(Tensor[] self, Scalar[] scalars) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_maximum(at::TensorList self, at::ArrayRef<at::Scalar> scalars) {
+    return at::_ops::_foreach_maximum_ScalarList::call(self, scalars);
+}
+
+// aten::_foreach_maximum_.ScalarList(Tensor(a!)[] self, Scalar[] scalars) -> ()
+inline void _foreach_maximum_(at::TensorList self, at::ArrayRef<at::Scalar> scalars) {
+    return at::_ops::_foreach_maximum__ScalarList::call(self, scalars);
+}
+
+// aten::_foreach_maximum.Scalar_out(Tensor[] self, Scalar scalar, *, Tensor(a!)[] out) -> ()
+inline void _foreach_maximum_out(at::TensorList out, at::TensorList self, const at::Scalar & scalar) {
+    return at::_ops::_foreach_maximum_Scalar_out::call(self, scalar, out);
+}
+// aten::_foreach_maximum.Scalar_out(Tensor[] self, Scalar scalar, *, Tensor(a!)[] out) -> ()
+inline void _foreach_maximum_outf(at::TensorList self, const at::Scalar & scalar, at::TensorList out) {
+    return at::_ops::_foreach_maximum_Scalar_out::call(self, scalar, out);
+}
+
+// aten::_foreach_maximum.List_out(Tensor[] self, Tensor[] other, *, Tensor(a!)[] out) -> ()
+inline void _foreach_maximum_out(at::TensorList out, at::TensorList self, at::TensorList other) {
+    return at::_ops::_foreach_maximum_List_out::call(self, other, out);
+}
+// aten::_foreach_maximum.List_out(Tensor[] self, Tensor[] other, *, Tensor(a!)[] out) -> ()
+inline void _foreach_maximum_outf(at::TensorList self, at::TensorList other, at::TensorList out) {
+    return at::_ops::_foreach_maximum_List_out::call(self, other, out);
+}
+
+// aten::_foreach_maximum.ScalarList_out(Tensor[] self, Scalar[] scalars, *, Tensor(a!)[] out) -> ()
+inline void _foreach_maximum_out(at::TensorList out, at::TensorList self, at::ArrayRef<at::Scalar> scalars) {
+    return at::_ops::_foreach_maximum_ScalarList_out::call(self, scalars, out);
+}
+// aten::_foreach_maximum.ScalarList_out(Tensor[] self, Scalar[] scalars, *, Tensor(a!)[] out) -> ()
+inline void _foreach_maximum_outf(at::TensorList self, at::ArrayRef<at::Scalar> scalars, at::TensorList out) {
+    return at::_ops::_foreach_maximum_ScalarList_out::call(self, scalars, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_minimum_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_minimum_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..40d0ff50a781983a5442ee428401a57dc04b4ae1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_minimum_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,39 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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::vector<at::Tensor> _foreach_minimum(at::TensorList self, const at::Scalar & scalar);
+TORCH_API void _foreach_minimum_out(at::TensorList out, at::TensorList self, const at::Scalar & scalar);
+TORCH_API void _foreach_minimum_outf(at::TensorList self, const at::Scalar & scalar, at::TensorList out);
+TORCH_API void _foreach_minimum_(at::TensorList self, const at::Scalar & scalar);
+TORCH_API ::std::vector<at::Tensor> _foreach_minimum(at::TensorList self, at::TensorList other);
+TORCH_API void _foreach_minimum_out(at::TensorList out, at::TensorList self, at::TensorList other);
+TORCH_API void _foreach_minimum_outf(at::TensorList self, at::TensorList other, at::TensorList out);
+TORCH_API void _foreach_minimum_(at::TensorList self, at::TensorList other);
+TORCH_API ::std::vector<at::Tensor> _foreach_minimum(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+TORCH_API void _foreach_minimum_out(at::TensorList out, at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+TORCH_API void _foreach_minimum_outf(at::TensorList self, at::ArrayRef<at::Scalar> scalars, at::TensorList out);
+TORCH_API void _foreach_minimum_(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_neg_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_neg_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..707bdf0760ef7719099995449f7b6fa0ec870600
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_neg_native.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_neg_slow(at::TensorList self);
+TORCH_API void _foreach_neg_out(at::TensorList self, at::TensorList out);
+TORCH_API void foreach_tensor_neg_slow_(at::TensorList self);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_neg_cuda(at::TensorList self);
+TORCH_API void foreach_tensor_neg_cuda_(at::TensorList self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_norm_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_norm_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..50ef46869442a9e2cfacbf96dee2106d22c65b5d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_norm_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_norm_slow(at::TensorList self, const at::Scalar & ord=2, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API void _foreach_norm_Scalar_out(at::TensorList self, const at::Scalar & ord, ::std::optional<at::ScalarType> dtype, at::TensorList out);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_norm_cuda(at::TensorList self, const at::Scalar & ord=2, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_pow_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_pow_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..468ca3f592d7e45203d2ceecd9a11551855e289a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_pow_cuda_dispatch.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API ::std::vector<at::Tensor> _foreach_pow(at::TensorList self, at::TensorList exponent);
+TORCH_API void _foreach_pow_(at::TensorList self, at::TensorList exponent);
+TORCH_API ::std::vector<at::Tensor> _foreach_pow(at::TensorList self, const at::Scalar & exponent);
+TORCH_API void _foreach_pow_(at::TensorList self, const at::Scalar & exponent);
+TORCH_API ::std::vector<at::Tensor> _foreach_pow(at::TensorList self, at::ArrayRef<at::Scalar> exponent);
+TORCH_API void _foreach_pow_(at::TensorList self, at::ArrayRef<at::Scalar> exponent);
+TORCH_API ::std::vector<at::Tensor> _foreach_pow(const at::Scalar & self, at::TensorList exponent);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_round_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_round_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..bc84887e27e0c66488596c7649126f1856788030
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_round_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_round {
+  using schema = ::std::vector<at::Tensor> (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_round";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_round(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_round_ {
+  using schema = void (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_round_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_round_(Tensor(a!)[] self) -> ()";
+  static void call(at::TensorList self);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_round_out {
+  using schema = void (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_round";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_foreach_round.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sigmoid_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sigmoid_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..21b5a2e0a8a88061e354cc3be3c35652411ad506
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sigmoid_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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::vector<at::Tensor> _foreach_sigmoid(at::TensorList self);
+TORCH_API void _foreach_sigmoid_out(at::TensorList out, at::TensorList self);
+TORCH_API void _foreach_sigmoid_outf(at::TensorList self, at::TensorList out);
+TORCH_API void _foreach_sigmoid_(at::TensorList self);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sigmoid_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sigmoid_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..58b3498ddcd7b476c2ee3732c7d72ed2dc60dad0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sigmoid_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_sigmoid {
+  using schema = ::std::vector<at::Tensor> (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sigmoid";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_sigmoid(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_sigmoid_ {
+  using schema = void (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sigmoid_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_sigmoid_(Tensor(a!)[] self) -> ()";
+  static void call(at::TensorList self);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_sigmoid_out {
+  using schema = void (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sigmoid";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_foreach_sigmoid.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sin_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sin_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..158359447f6fb965c11ffa7be101831ed32b7ad7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sin_cuda_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API ::std::vector<at::Tensor> _foreach_sin(at::TensorList self);
+TORCH_API void _foreach_sin_(at::TensorList self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sinh.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sinh.h
new file mode 100644
index 0000000000000000000000000000000000000000..b417d5e260b666be35d667873c6c2268752f931c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sinh.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_foreach_sinh_ops.h>
+
+namespace at {
+
+
+// aten::_foreach_sinh(Tensor[] self) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_sinh(at::TensorList self) {
+    return at::_ops::_foreach_sinh::call(self);
+}
+
+// aten::_foreach_sinh_(Tensor(a!)[] self) -> ()
+inline void _foreach_sinh_(at::TensorList self) {
+    return at::_ops::_foreach_sinh_::call(self);
+}
+
+// aten::_foreach_sinh.out(Tensor[] self, *, Tensor(a!)[] out) -> ()
+inline void _foreach_sinh_out(at::TensorList out, at::TensorList self) {
+    return at::_ops::_foreach_sinh_out::call(self, out);
+}
+// aten::_foreach_sinh.out(Tensor[] self, *, Tensor(a!)[] out) -> ()
+inline void _foreach_sinh_outf(at::TensorList self, at::TensorList out) {
+    return at::_ops::_foreach_sinh_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sqrt_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sqrt_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..cec9f2125109bc25555fb550dde41f1ae0c4c337
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sqrt_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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::vector<at::Tensor> _foreach_sqrt(at::TensorList self);
+TORCH_API void _foreach_sqrt_out(at::TensorList out, at::TensorList self);
+TORCH_API void _foreach_sqrt_outf(at::TensorList self, at::TensorList out);
+TORCH_API void _foreach_sqrt_(at::TensorList self);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sub_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sub_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..309e4e16a89d9431a21db3ce73923d8b947ce90f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_sub_ops.h
@@ -0,0 +1,122 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_sub_Scalar {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sub";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "_foreach_sub.Scalar(Tensor[] self, Scalar scalar) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList self, const at::Scalar & scalar);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, const at::Scalar & scalar);
+};
+
+struct TORCH_API _foreach_sub__Scalar {
+  using schema = void (at::TensorList, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sub_";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "_foreach_sub_.Scalar(Tensor(a!)[] self, Scalar scalar) -> ()";
+  static void call(at::TensorList self, const at::Scalar & scalar);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, const at::Scalar & scalar);
+};
+
+struct TORCH_API _foreach_sub_List {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, at::TensorList, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sub";
+  static constexpr const char* overload_name = "List";
+  static constexpr const char* schema_str = "_foreach_sub.List(Tensor[] self, Tensor[] other, *, Scalar alpha=1) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList self, at::TensorList other, const at::Scalar & alpha);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList other, const at::Scalar & alpha);
+};
+
+struct TORCH_API _foreach_sub__List {
+  using schema = void (at::TensorList, at::TensorList, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sub_";
+  static constexpr const char* overload_name = "List";
+  static constexpr const char* schema_str = "_foreach_sub_.List(Tensor(a!)[] self, Tensor[] other, *, Scalar alpha=1) -> ()";
+  static void call(at::TensorList self, at::TensorList other, const at::Scalar & alpha);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList other, const at::Scalar & alpha);
+};
+
+struct TORCH_API _foreach_sub_ScalarList {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, at::ArrayRef<at::Scalar>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sub";
+  static constexpr const char* overload_name = "ScalarList";
+  static constexpr const char* schema_str = "_foreach_sub.ScalarList(Tensor[] self, Scalar[] scalars) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+};
+
+struct TORCH_API _foreach_sub__ScalarList {
+  using schema = void (at::TensorList, at::ArrayRef<at::Scalar>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sub_";
+  static constexpr const char* overload_name = "ScalarList";
+  static constexpr const char* schema_str = "_foreach_sub_.ScalarList(Tensor(a!)[] self, Scalar[] scalars) -> ()";
+  static void call(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+};
+
+struct TORCH_API _foreach_sub_Scalar_out {
+  using schema = void (at::TensorList, const at::Scalar &, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sub";
+  static constexpr const char* overload_name = "Scalar_out";
+  static constexpr const char* schema_str = "_foreach_sub.Scalar_out(Tensor[] self, Scalar scalar, *, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList self, const at::Scalar & scalar, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, const at::Scalar & scalar, at::TensorList out);
+};
+
+struct TORCH_API _foreach_sub_List_out {
+  using schema = void (at::TensorList, at::TensorList, const at::Scalar &, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sub";
+  static constexpr const char* overload_name = "List_out";
+  static constexpr const char* schema_str = "_foreach_sub.List_out(Tensor[] self, Tensor[] other, *, Scalar alpha=1, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList self, at::TensorList other, const at::Scalar & alpha, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList other, const at::Scalar & alpha, at::TensorList out);
+};
+
+struct TORCH_API _foreach_sub_ScalarList_out {
+  using schema = void (at::TensorList, at::ArrayRef<at::Scalar>, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_sub";
+  static constexpr const char* overload_name = "ScalarList_out";
+  static constexpr const char* schema_str = "_foreach_sub.ScalarList_out(Tensor[] self, Scalar[] scalars, *, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList self, at::ArrayRef<at::Scalar> scalars, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::ArrayRef<at::Scalar> scalars, at::TensorList out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_tan.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_tan.h
new file mode 100644
index 0000000000000000000000000000000000000000..985b48ac272d678734e51ed2a89bda5de6747c3f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_tan.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_foreach_tan_ops.h>
+
+namespace at {
+
+
+// aten::_foreach_tan(Tensor[] self) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_tan(at::TensorList self) {
+    return at::_ops::_foreach_tan::call(self);
+}
+
+// aten::_foreach_tan_(Tensor(a!)[] self) -> ()
+inline void _foreach_tan_(at::TensorList self) {
+    return at::_ops::_foreach_tan_::call(self);
+}
+
+// aten::_foreach_tan.out(Tensor[] self, *, Tensor(a!)[] out) -> ()
+inline void _foreach_tan_out(at::TensorList out, at::TensorList self) {
+    return at::_ops::_foreach_tan_out::call(self, out);
+}
+// aten::_foreach_tan.out(Tensor[] self, *, Tensor(a!)[] out) -> ()
+inline void _foreach_tan_outf(at::TensorList self, at::TensorList out) {
+    return at::_ops::_foreach_tan_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_tan_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_tan_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..c44d06063e0127ddb5eb6b5c521840f4484ec503
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_tan_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_tan {
+  using schema = ::std::vector<at::Tensor> (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_tan";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_tan(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_tan_ {
+  using schema = void (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_tan_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_foreach_tan_(Tensor(a!)[] self) -> ()";
+  static void call(at::TensorList self);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self);
+};
+
+struct TORCH_API _foreach_tan_out {
+  using schema = void (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_foreach_tan";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_foreach_tan.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_tanh_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_tanh_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..02930b3ab45a704a44825a4ac0f2993f45075010
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_foreach_tanh_cuda_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API ::std::vector<at::Tensor> _foreach_tanh(at::TensorList self);
+TORCH_API void _foreach_tanh_(at::TensorList self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_async_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_async_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..396406e74d4c64ae0201b7d928fa6b3afa2fcfae
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_async_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _functional_assert_async_msg_cpu(const at::Tensor & self, c10::string_view assert_msg, const at::Tensor & dep_token);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_async_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_async_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..ad4e53e4898b97553362eef909aa89ce236a34ce
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_async_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _functional_assert_async_msg {
+  using schema = at::Tensor (const at::Tensor &, c10::string_view, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_functional_assert_async";
+  static constexpr const char* overload_name = "msg";
+  static constexpr const char* schema_str = "_functional_assert_async.msg(Tensor self, str assert_msg, Tensor dep_token) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::string_view assert_msg, const at::Tensor & dep_token);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::string_view assert_msg, const at::Tensor & dep_token);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_scalar.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_scalar.h
new file mode 100644
index 0000000000000000000000000000000000000000..0166eb7ebb9f3dbacad263df812f11d5b88eed6b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_scalar.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_functional_assert_scalar_ops.h>
+
+namespace at {
+
+
+// aten::_functional_assert_scalar(Scalar self, str assert_msg, Tensor dep_token) -> Tensor
+inline at::Tensor _functional_assert_scalar(const at::Scalar & self, c10::string_view assert_msg, const at::Tensor & dep_token) {
+    return at::_ops::_functional_assert_scalar::call(self, assert_msg, dep_token);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_scalar_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_scalar_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b043faee8618cd4873bbf2dec283d7a6e2322da3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_functional_assert_scalar_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _functional_assert_scalar(const at::Scalar & self, c10::string_view assert_msg, const at::Tensor & dep_token);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_adamw_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_adamw_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..04b8be9f3b72af3e65c659f31e373f4d5d0e1f2b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_adamw_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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<::std::vector<at::Tensor>,::std::vector<at::Tensor>,::std::vector<at::Tensor>,::std::vector<at::Tensor>,::std::vector<at::Tensor>> _fused_adamw(at::TensorList self, at::TensorList grads, at::TensorList exp_avgs, at::TensorList exp_avg_sqs, at::TensorList max_exp_avg_sqs, at::TensorList state_steps, double lr, double beta1, double beta2, double weight_decay, double eps, bool amsgrad, bool maximize, const ::std::optional<at::Tensor> & grad_scale={}, const ::std::optional<at::Tensor> & found_inf={});
+TORCH_API void _fused_adamw_out(at::TensorList out, at::TensorList self, at::TensorList grads, at::TensorList exp_avgs, at::TensorList exp_avg_sqs, at::TensorList max_exp_avg_sqs, at::TensorList state_steps, double lr, double beta1, double beta2, double weight_decay, double eps, bool amsgrad, bool maximize, const ::std::optional<at::Tensor> & grad_scale={}, const ::std::optional<at::Tensor> & found_inf={});
+TORCH_API void _fused_adamw_outf(at::TensorList self, at::TensorList grads, at::TensorList exp_avgs, at::TensorList exp_avg_sqs, at::TensorList max_exp_avg_sqs, at::TensorList state_steps, double lr, double beta1, double beta2, double weight_decay, double eps, bool amsgrad, bool maximize, const ::std::optional<at::Tensor> & grad_scale, const ::std::optional<at::Tensor> & found_inf, at::TensorList out);
+TORCH_API ::std::tuple<::std::vector<at::Tensor>,::std::vector<at::Tensor>,::std::vector<at::Tensor>,::std::vector<at::Tensor>,::std::vector<at::Tensor>> _fused_adamw(at::TensorList self, at::TensorList grads, at::TensorList exp_avgs, at::TensorList exp_avg_sqs, at::TensorList max_exp_avg_sqs, at::TensorList state_steps, const at::Tensor & lr, double beta1, double beta2, double weight_decay, double eps, bool amsgrad, bool maximize, const ::std::optional<at::Tensor> & grad_scale={}, const ::std::optional<at::Tensor> & found_inf={});
+TORCH_API void _fused_adamw_out(at::TensorList out, at::TensorList self, at::TensorList grads, at::TensorList exp_avgs, at::TensorList exp_avg_sqs, at::TensorList max_exp_avg_sqs, at::TensorList state_steps, const at::Tensor & lr, double beta1, double beta2, double weight_decay, double eps, bool amsgrad, bool maximize, const ::std::optional<at::Tensor> & grad_scale={}, const ::std::optional<at::Tensor> & found_inf={});
+TORCH_API void _fused_adamw_outf(at::TensorList self, at::TensorList grads, at::TensorList exp_avgs, at::TensorList exp_avg_sqs, at::TensorList max_exp_avg_sqs, at::TensorList state_steps, const at::Tensor & lr, double beta1, double beta2, double weight_decay, double eps, bool amsgrad, bool maximize, const ::std::optional<at::Tensor> & grad_scale, const ::std::optional<at::Tensor> & found_inf, at::TensorList out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_dropout_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_dropout_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..eef55b71093157cfe20178dfcd977fa016c6b344
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_dropout_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> _fused_dropout_out(const at::Tensor & self, double p, ::std::optional<at::Generator> generator, at::Tensor & out0, at::Tensor & out1);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> fused_dropout_cuda(const at::Tensor & self, double p, ::std::optional<at::Generator> generator=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_dropout_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_dropout_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..b8edfbba146a5511959be2ac49d0e24f72e26293
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_dropout_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _fused_dropout {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, double, ::std::optional<at::Generator>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_fused_dropout";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_fused_dropout(Tensor self, float p, Generator? generator=None) -> (Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, double p, ::std::optional<at::Generator> generator);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, double p, ::std::optional<at::Generator> generator);
+};
+
+struct TORCH_API _fused_dropout_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, double, ::std::optional<at::Generator>, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_fused_dropout";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_fused_dropout.out(Tensor self, float p, Generator? generator=None, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, double p, ::std::optional<at::Generator> generator, at::Tensor & out0, at::Tensor & out1);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, double p, ::std::optional<at::Generator> generator, at::Tensor & out0, at::Tensor & out1);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_rms_norm_backward.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_rms_norm_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..e68188a1597dfb196f85ed77c5bf054674217adb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_rms_norm_backward.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_fused_rms_norm_backward_ops.h>
+
+namespace at {
+
+
+// aten::_fused_rms_norm_backward(Tensor grad_out, Tensor input, int[] normalized_shape, Tensor rstd, Tensor? weight, bool[2] output_mask) -> (Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor> _fused_rms_norm_backward(const at::Tensor & grad_out, const at::Tensor & input, at::IntArrayRef normalized_shape, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, ::std::array<bool,2> output_mask) {
+    return at::_ops::_fused_rms_norm_backward::call(grad_out, input, normalized_shape, rstd, weight, output_mask);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_sdp_choice_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_sdp_choice_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..02d3233cb1401769d7656da7e3ec5a04fe1a06b6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_sdp_choice_meta_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 int64_t _fused_sdp_choice(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const ::std::optional<at::Tensor> & attn_mask={}, double dropout_p=0.0, bool is_causal=false, ::std::optional<double> scale=::std::nullopt, bool enable_gqa=false);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_sgd.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_sgd.h
new file mode 100644
index 0000000000000000000000000000000000000000..730d1b8fb7395d274c380ed3e005029f60b1cab3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fused_sgd.h
@@ -0,0 +1,69 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_fused_sgd_ops.h>
+
+namespace at {
+
+
+// aten::_fused_sgd_(Tensor(a!)[] self, Tensor(b!)[] grads, Tensor(c!)[] momentum_buffer_list, *, float weight_decay, float momentum, float lr, float dampening, bool nesterov, bool maximize, bool is_first_step, Tensor? grad_scale=None, Tensor? found_inf=None) -> ()
+inline void _fused_sgd_(at::TensorList self, at::TensorList grads, at::TensorList momentum_buffer_list, double weight_decay, double momentum, double lr, double dampening, bool nesterov, bool maximize, bool is_first_step, const ::std::optional<at::Tensor> & grad_scale={}, const ::std::optional<at::Tensor> & found_inf={}) {
+    return at::_ops::_fused_sgd_::call(self, grads, momentum_buffer_list, weight_decay, momentum, lr, dampening, nesterov, maximize, is_first_step, grad_scale, found_inf);
+}
+
+// aten::_fused_sgd_.tensor_lr(Tensor(a!)[] self, Tensor(b!)[] grads, Tensor(c!)[] momentum_buffer_list, *, float weight_decay, float momentum, Tensor lr, float dampening, bool nesterov, bool maximize, bool is_first_step, Tensor? grad_scale=None, Tensor? found_inf=None) -> ()
+inline void _fused_sgd_(at::TensorList self, at::TensorList grads, at::TensorList momentum_buffer_list, double weight_decay, double momentum, const at::Tensor & lr, double dampening, bool nesterov, bool maximize, bool is_first_step, const ::std::optional<at::Tensor> & grad_scale={}, const ::std::optional<at::Tensor> & found_inf={}) {
+    return at::_ops::_fused_sgd__tensor_lr::call(self, grads, momentum_buffer_list, weight_decay, momentum, lr, dampening, nesterov, maximize, is_first_step, grad_scale, found_inf);
+}
+
+// aten::_fused_sgd.out(Tensor[] self, Tensor(b!)[] grads, Tensor(c!)[] momentum_buffer_list, *, float weight_decay, float momentum, float lr, float dampening, bool nesterov, bool maximize, bool is_first_step, Tensor? grad_scale=None, Tensor? found_inf=None, Tensor(a!)[] out) -> ()
+inline void _fused_sgd_out(at::TensorList out, at::TensorList self, at::TensorList grads, at::TensorList momentum_buffer_list, double weight_decay, double momentum, double lr, double dampening, bool nesterov, bool maximize, bool is_first_step, const ::std::optional<at::Tensor> & grad_scale={}, const ::std::optional<at::Tensor> & found_inf={}) {
+    return at::_ops::_fused_sgd_out::call(self, grads, momentum_buffer_list, weight_decay, momentum, lr, dampening, nesterov, maximize, is_first_step, grad_scale, found_inf, out);
+}
+// aten::_fused_sgd.out(Tensor[] self, Tensor(b!)[] grads, Tensor(c!)[] momentum_buffer_list, *, float weight_decay, float momentum, float lr, float dampening, bool nesterov, bool maximize, bool is_first_step, Tensor? grad_scale=None, Tensor? found_inf=None, Tensor(a!)[] out) -> ()
+inline void _fused_sgd_outf(at::TensorList self, at::TensorList grads, at::TensorList momentum_buffer_list, double weight_decay, double momentum, double lr, double dampening, bool nesterov, bool maximize, bool is_first_step, const ::std::optional<at::Tensor> & grad_scale, const ::std::optional<at::Tensor> & found_inf, at::TensorList out) {
+    return at::_ops::_fused_sgd_out::call(self, grads, momentum_buffer_list, weight_decay, momentum, lr, dampening, nesterov, maximize, is_first_step, grad_scale, found_inf, out);
+}
+
+// aten::_fused_sgd(Tensor[] self, Tensor[] grads, Tensor[] momentum_buffer_list, *, float weight_decay, float momentum, float lr, float dampening, bool nesterov, bool maximize, bool is_first_step, Tensor? grad_scale=None, Tensor? found_inf=None) -> (Tensor[] self_out, Tensor[] grads_out, Tensor[] momentum_buffer_list_out)
+inline ::std::tuple<::std::vector<at::Tensor>,::std::vector<at::Tensor>,::std::vector<at::Tensor>> _fused_sgd(at::TensorList self, at::TensorList grads, at::TensorList momentum_buffer_list, double weight_decay, double momentum, double lr, double dampening, bool nesterov, bool maximize, bool is_first_step, const ::std::optional<at::Tensor> & grad_scale={}, const ::std::optional<at::Tensor> & found_inf={}) {
+    return at::_ops::_fused_sgd::call(self, grads, momentum_buffer_list, weight_decay, momentum, lr, dampening, nesterov, maximize, is_first_step, grad_scale, found_inf);
+}
+
+// aten::_fused_sgd.tensor_lr_out(Tensor[] self, Tensor(b!)[] grads, Tensor(c!)[] momentum_buffer_list, *, float weight_decay, float momentum, Tensor lr, float dampening, bool nesterov, bool maximize, bool is_first_step, Tensor? grad_scale=None, Tensor? found_inf=None, Tensor(a!)[] out) -> ()
+inline void _fused_sgd_out(at::TensorList out, at::TensorList self, at::TensorList grads, at::TensorList momentum_buffer_list, double weight_decay, double momentum, const at::Tensor & lr, double dampening, bool nesterov, bool maximize, bool is_first_step, const ::std::optional<at::Tensor> & grad_scale={}, const ::std::optional<at::Tensor> & found_inf={}) {
+    return at::_ops::_fused_sgd_tensor_lr_out::call(self, grads, momentum_buffer_list, weight_decay, momentum, lr, dampening, nesterov, maximize, is_first_step, grad_scale, found_inf, out);
+}
+// aten::_fused_sgd.tensor_lr_out(Tensor[] self, Tensor(b!)[] grads, Tensor(c!)[] momentum_buffer_list, *, float weight_decay, float momentum, Tensor lr, float dampening, bool nesterov, bool maximize, bool is_first_step, Tensor? grad_scale=None, Tensor? found_inf=None, Tensor(a!)[] out) -> ()
+inline void _fused_sgd_outf(at::TensorList self, at::TensorList grads, at::TensorList momentum_buffer_list, double weight_decay, double momentum, const at::Tensor & lr, double dampening, bool nesterov, bool maximize, bool is_first_step, const ::std::optional<at::Tensor> & grad_scale, const ::std::optional<at::Tensor> & found_inf, at::TensorList out) {
+    return at::_ops::_fused_sgd_tensor_lr_out::call(self, grads, momentum_buffer_list, weight_decay, momentum, lr, dampening, nesterov, maximize, is_first_step, grad_scale, found_inf, out);
+}
+
+// aten::_fused_sgd.tensor_lr(Tensor[] self, Tensor[] grads, Tensor[] momentum_buffer_list, *, float weight_decay, float momentum, Tensor lr, float dampening, bool nesterov, bool maximize, bool is_first_step, Tensor? grad_scale=None, Tensor? found_inf=None) -> (Tensor[] self_out, Tensor[] grads_out, Tensor[] momentum_buffer_list_out)
+inline ::std::tuple<::std::vector<at::Tensor>,::std::vector<at::Tensor>,::std::vector<at::Tensor>> _fused_sgd(at::TensorList self, at::TensorList grads, at::TensorList momentum_buffer_list, double weight_decay, double momentum, const at::Tensor & lr, double dampening, bool nesterov, bool maximize, bool is_first_step, const ::std::optional<at::Tensor> & grad_scale={}, const ::std::optional<at::Tensor> & found_inf={}) {
+    return at::_ops::_fused_sgd_tensor_lr::call(self, grads, momentum_buffer_list, weight_decay, momentum, lr, dampening, nesterov, maximize, is_first_step, grad_scale, found_inf);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fw_primal_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fw_primal_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..67a5dd7e25bdbff4a8fcb3f56fc53c467e068adc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_fw_primal_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _fw_primal(const at::Tensor & self, int64_t level);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_grid_sampler_2d_cpu_fallback_backward_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_grid_sampler_2d_cpu_fallback_backward_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..96144cae319b0bede01ea0dbebc38fb9dd428cae
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_grid_sampler_2d_cpu_fallback_backward_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,at::Tensor> _grid_sampler_2d_cpu_fallback_backward(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_grid_sampler_2d_cpu_fallback_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_grid_sampler_2d_cpu_fallback_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..68353e05715333e03fdbef227300a7abdd0dcd0b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_grid_sampler_2d_cpu_fallback_backward_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> _grid_sampler_2d_cpu_fallback_backward(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_grouped_mm_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_grouped_mm_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..7b5085930c0d05d3019e250feec8a9dee5a07648
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_grouped_mm_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _grouped_mm {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, ::std::optional<at::ScalarType>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_grouped_mm";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_grouped_mm(Tensor self, Tensor mat2, Tensor? offs=None, Tensor? bias=None, ScalarType? out_dtype=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & mat2, const ::std::optional<at::Tensor> & offs, const ::std::optional<at::Tensor> & bias, ::std::optional<at::ScalarType> out_dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat2, const ::std::optional<at::Tensor> & offs, const ::std::optional<at::Tensor> & bias, ::std::optional<at::ScalarType> out_dtype);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_has_compatible_shallow_copy_type_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_has_compatible_shallow_copy_type_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..0c5467659e037beca65e47de47735d752c93058e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_has_compatible_shallow_copy_type_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _has_compatible_shallow_copy_type {
+  using schema = bool (const at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_has_compatible_shallow_copy_type";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_has_compatible_shallow_copy_type(Tensor self, Tensor from) -> bool";
+  static bool call(const at::Tensor & self, const at::Tensor & from);
+  static bool redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & from);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_has_same_storage_numel_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_has_same_storage_numel_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..20a0af13aa410409b21edd06aa631d3b4ff06510
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_has_same_storage_numel_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API bool _has_same_storage_numel(const at::Tensor & self, const at::Tensor & other);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_has_same_storage_numel_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_has_same_storage_numel_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8ca657a63340a5eab3d95f83771673eafe9de118
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_has_same_storage_numel_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _has_same_storage_numel {
+  using schema = bool (const at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_has_same_storage_numel";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_has_same_storage_numel(Tensor self, Tensor other) -> bool";
+  static bool call(const at::Tensor & self, const at::Tensor & other);
+  static bool redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_histogramdd_bin_edges_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_histogramdd_bin_edges_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3ebbd414f8ae5436b41d69a36fa021bc9d7b5993
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_histogramdd_bin_edges_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API ::std::vector<at::Tensor> _histogramdd_bin_edges(const at::Tensor & self, at::IntArrayRef bins, ::std::optional<at::ArrayRef<double>> range=::std::nullopt, const ::std::optional<at::Tensor> & weight={}, bool density=false);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_histogramdd_from_bin_tensors_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_histogramdd_from_bin_tensors_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ec18f89f6174bfe202e7fad442c7e3f2fb5b9da9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_histogramdd_from_bin_tensors_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _histogramdd_from_bin_tensors(const at::Tensor & self, at::TensorList bins, const ::std::optional<at::Tensor> & weight={}, bool density=false);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_index_put_impl_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_index_put_impl_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3c43dafd20205031389d9ef3ba757490d5acfc57
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_index_put_impl_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _index_put_impl(const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate=false, bool unsafe=false);
+TORCH_API at::Tensor & _index_put_impl_out(at::Tensor & out, const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate=false, bool unsafe=false);
+TORCH_API at::Tensor & _index_put_impl_outf(const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate, bool unsafe, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_indices_copy_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_indices_copy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..d5005c6e00b76f6eb410c618dcf36944cfc45ab2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_indices_copy_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _indices_copy_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor _indices_copy(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_int_mm.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_int_mm.h
new file mode 100644
index 0000000000000000000000000000000000000000..6884a3515ee64cf41e0a86f61e10df09c3ef0b09
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_int_mm.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_int_mm_ops.h>
+
+namespace at {
+
+
+// aten::_int_mm(Tensor self, Tensor mat2) -> Tensor
+inline at::Tensor _int_mm(const at::Tensor & self, const at::Tensor & mat2) {
+    return at::_ops::_int_mm::call(self, mat2);
+}
+
+// aten::_int_mm.out(Tensor self, Tensor mat2, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _int_mm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat2) {
+    return at::_ops::_int_mm_out::call(self, mat2, out);
+}
+// aten::_int_mm.out(Tensor self, Tensor mat2, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _int_mm_outf(const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out) {
+    return at::_ops::_int_mm_out::call(self, mat2, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_det.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_det.h
new file mode 100644
index 0000000000000000000000000000000000000000..b7adc984801ed7505db6dd0c0d77f2506200579c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_det.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_linalg_det_ops.h>
+
+namespace at {
+
+
+// aten::_linalg_det(Tensor A) -> (Tensor result, Tensor LU, Tensor pivots)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _linalg_det(const at::Tensor & A) {
+    return at::_ops::_linalg_det::call(A);
+}
+
+// aten::_linalg_det.result(Tensor A, *, Tensor(a!) result, Tensor(b!) LU, Tensor(c!) pivots) -> (Tensor(a!) result, Tensor(b!) LU, Tensor(c!) pivots)
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _linalg_det_out(at::Tensor & result, at::Tensor & LU, at::Tensor & pivots, const at::Tensor & A) {
+    return at::_ops::_linalg_det_result::call(A, result, LU, pivots);
+}
+// aten::_linalg_det.result(Tensor A, *, Tensor(a!) result, Tensor(b!) LU, Tensor(c!) pivots) -> (Tensor(a!) result, Tensor(b!) LU, Tensor(c!) pivots)
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _linalg_det_outf(const at::Tensor & A, at::Tensor & result, at::Tensor & LU, at::Tensor & pivots) {
+    return at::_ops::_linalg_det_result::call(A, result, LU, pivots);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_det_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_det_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..036498f838cf966012d6bfb9c576ae4352ff7781
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_det_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _linalg_det(const at::Tensor & A);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _linalg_det_out(at::Tensor & result, at::Tensor & LU, at::Tensor & pivots, const at::Tensor & A);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _linalg_det_outf(const at::Tensor & A, at::Tensor & result, at::Tensor & LU, at::Tensor & pivots);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_eigh_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_eigh_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..eb81dac129b723aa8e4b4a8e972ded1c25150f67
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_eigh_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/_linalg_eigh_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured__linalg_eigh_out : public at::meta::structured__linalg_eigh {
+void impl(const at::Tensor & A, c10::string_view UPLO, bool compute_v, const at::Tensor & eigenvalues, const at::Tensor & eigenvectors);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_eigvals.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_eigvals.h
new file mode 100644
index 0000000000000000000000000000000000000000..527d09d1f21639fd529b0f2d56dab03d6bdbba60
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_eigvals.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_linalg_eigvals_ops.h>
+
+namespace at {
+
+
+// aten::_linalg_eigvals(Tensor self) -> Tensor
+inline at::Tensor _linalg_eigvals(const at::Tensor & self) {
+    return at::_ops::_linalg_eigvals::call(self);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_eigvals_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_eigvals_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..4bc3a3b5b15e2b25c3cccdffe709e7dabe28dc72
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_eigvals_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _linalg_eigvals {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_linalg_eigvals";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_linalg_eigvals(Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_slogdet_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_slogdet_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..66536b63a37e900008ffdb612e3babed97bdde80
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_slogdet_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> _linalg_slogdet(const at::Tensor & A);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> _linalg_slogdet_out(at::Tensor & sign, at::Tensor & logabsdet, at::Tensor & LU, at::Tensor & pivots, const at::Tensor & A);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> _linalg_slogdet_outf(const at::Tensor & A, at::Tensor & sign, at::Tensor & logabsdet, at::Tensor & LU, at::Tensor & pivots);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_solve_ex_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_solve_ex_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..27faa7d77775d9a567131c87d8ade5ac6b54d0e2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_solve_ex_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> _linalg_solve_ex(const at::Tensor & A, const at::Tensor & B, bool left=true, bool check_errors=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> _linalg_solve_ex_out(at::Tensor & result, at::Tensor & LU, at::Tensor & pivots, at::Tensor & info, const at::Tensor & A, const at::Tensor & B, bool left=true, bool check_errors=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> _linalg_solve_ex_outf(const at::Tensor & A, const at::Tensor & B, bool left, bool check_errors, at::Tensor & result, at::Tensor & LU, at::Tensor & pivots, at::Tensor & info);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_svd_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_svd_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..667906125fd20d61f936f2a06114eb51d0cc7c0a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_svd_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _linalg_svd(const at::Tensor & A, bool full_matrices=false, bool compute_uv=true, ::std::optional<c10::string_view> driver=::std::nullopt);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _linalg_svd_out(at::Tensor & U, at::Tensor & S, at::Tensor & Vh, const at::Tensor & A, bool full_matrices=false, bool compute_uv=true, ::std::optional<c10::string_view> driver=::std::nullopt);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _linalg_svd_outf(const at::Tensor & A, bool full_matrices, bool compute_uv, ::std::optional<c10::string_view> driver, at::Tensor & U, at::Tensor & S, at::Tensor & Vh);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_svd_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_svd_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a4707911e4f9f43143e7d3f2874a2d6b637ca73c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_svd_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> _linalg_svd(const at::Tensor & A, bool full_matrices=false, bool compute_uv=true, ::std::optional<c10::string_view> driver=::std::nullopt);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _linalg_svd_out(at::Tensor & U, at::Tensor & S, at::Tensor & Vh, const at::Tensor & A, bool full_matrices=false, bool compute_uv=true, ::std::optional<c10::string_view> driver=::std::nullopt);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _linalg_svd_outf(const at::Tensor & A, bool full_matrices, bool compute_uv, ::std::optional<c10::string_view> driver, at::Tensor & U, at::Tensor & S, at::Tensor & Vh);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_svd_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_svd_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..27b1a2bee72c1ae1f92c3d5fe14303605094ee5c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_linalg_svd_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/_linalg_svd_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured__linalg_svd_out : public at::meta::structured__linalg_svd {
+void impl(const at::Tensor & A, bool full_matrices, bool compute_uv, ::std::optional<c10::string_view> driver, const at::Tensor & U, const at::Tensor & S, const at::Tensor & Vh);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_log_softmax_backward_data_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_log_softmax_backward_data_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c572c56100c3aa0d970729558f101ffb8228fbd3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_log_softmax_backward_data_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _log_softmax_backward_data(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype);
+TORCH_API at::Tensor & _log_softmax_backward_data_out(at::Tensor & out, const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype);
+TORCH_API at::Tensor & _log_softmax_backward_data_outf(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_logcumsumexp_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_logcumsumexp_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..13442417e38ef45875e9da399ac74601c018ea66
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_logcumsumexp_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _logcumsumexp(const at::Tensor & self, int64_t dim);
+TORCH_API at::Tensor & _logcumsumexp_out(at::Tensor & out, const at::Tensor & self, int64_t dim);
+TORCH_API at::Tensor & _logcumsumexp_outf(const at::Tensor & self, int64_t dim, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_lu_with_info.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_lu_with_info.h
new file mode 100644
index 0000000000000000000000000000000000000000..907ff3f14c12147356a55dad34809c865d4f0e47
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_lu_with_info.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_lu_with_info_ops.h>
+
+namespace at {
+
+
+// aten::_lu_with_info(Tensor self, bool pivot=True, bool check_errors=True) -> (Tensor LU, Tensor pivots, Tensor info)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _lu_with_info(const at::Tensor & self, bool pivot=true, bool check_errors=true) {
+    return at::_ops::_lu_with_info::call(self, pivot, check_errors);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_make_dual_copy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_make_dual_copy.h
new file mode 100644
index 0000000000000000000000000000000000000000..de5a7f1b1edb8c32f6fad11106872569a5cf77a4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_make_dual_copy.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_make_dual_copy_ops.h>
+
+namespace at {
+
+
+// aten::_make_dual_copy(Tensor primal, Tensor tangent, int level) -> Tensor
+inline at::Tensor _make_dual_copy(const at::Tensor & primal, const at::Tensor & tangent, int64_t level) {
+    return at::_ops::_make_dual_copy::call(primal, tangent, level);
+}
+
+// aten::_make_dual_copy.out(Tensor primal, Tensor tangent, int level, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _make_dual_copy_out(at::Tensor & out, const at::Tensor & primal, const at::Tensor & tangent, int64_t level) {
+    return at::_ops::_make_dual_copy_out::call(primal, tangent, level, out);
+}
+// aten::_make_dual_copy.out(Tensor primal, Tensor tangent, int level, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _make_dual_copy_outf(const at::Tensor & primal, const at::Tensor & tangent, int64_t level, at::Tensor & out) {
+    return at::_ops::_make_dual_copy_out::call(primal, tangent, level, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_make_dual_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_make_dual_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..acaef91e8f2278bf6ebe550c93d27679f9202019
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_make_dual_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _make_dual {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_make_dual";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_make_dual(Tensor(a) primal, Tensor tangent, int level) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & primal, const at::Tensor & tangent, int64_t level);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & primal, const at::Tensor & tangent, int64_t level);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_make_per_channel_quantized_tensor_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_make_per_channel_quantized_tensor_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..513e857ea0d24ec93d4b26f4ee79dd35d913ef16
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_make_per_channel_quantized_tensor_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _make_per_channel_quantized_tensor {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_make_per_channel_quantized_tensor";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_make_per_channel_quantized_tensor(Tensor self, Tensor scale, Tensor zero_point, int axis) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis);
+};
+
+struct TORCH_API _make_per_channel_quantized_tensor_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, const at::Tensor &, int64_t, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_make_per_channel_quantized_tensor";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_make_per_channel_quantized_tensor.out(Tensor self, Tensor scale, Tensor zero_point, int axis, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_masked_softmax.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_masked_softmax.h
new file mode 100644
index 0000000000000000000000000000000000000000..56ce5e4d34b49b56e19e160932e13279612528bc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_masked_softmax.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_masked_softmax_ops.h>
+
+namespace at {
+
+
+// aten::_masked_softmax(Tensor self, Tensor mask, int? dim=None, int? mask_type=None) -> Tensor
+inline at::Tensor _masked_softmax(const at::Tensor & self, const at::Tensor & mask, ::std::optional<int64_t> dim=::std::nullopt, ::std::optional<int64_t> mask_type=::std::nullopt) {
+    return at::_ops::_masked_softmax::call(self, mask, dim, mask_type);
+}
+
+// aten::_masked_softmax.out(Tensor self, Tensor mask, int? dim=None, int? mask_type=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _masked_softmax_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mask, ::std::optional<int64_t> dim=::std::nullopt, ::std::optional<int64_t> mask_type=::std::nullopt) {
+    return at::_ops::_masked_softmax_out::call(self, mask, dim, mask_type, out);
+}
+// aten::_masked_softmax.out(Tensor self, Tensor mask, int? dim=None, int? mask_type=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _masked_softmax_outf(const at::Tensor & self, const at::Tensor & mask, ::std::optional<int64_t> dim, ::std::optional<int64_t> mask_type, at::Tensor & out) {
+    return at::_ops::_masked_softmax_out::call(self, mask, dim, mask_type, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_masked_softmax_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_masked_softmax_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..59c3453ecc8d7b6facd3268e2805cbd6d84f5f71
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_masked_softmax_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _masked_softmax_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mask, ::std::optional<int64_t> dim=::std::nullopt, ::std::optional<int64_t> mask_type=::std::nullopt);
+TORCH_API at::Tensor & _masked_softmax_outf(const at::Tensor & self, const at::Tensor & mask, ::std::optional<int64_t> dim, ::std::optional<int64_t> mask_type, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_masked_softmax_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_masked_softmax_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..cc92d5afde0f877330838534dd0c377adb92e376
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_masked_softmax_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _masked_softmax {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, ::std::optional<int64_t>, ::std::optional<int64_t>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_masked_softmax";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_masked_softmax(Tensor self, Tensor mask, int? dim=None, int? mask_type=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & mask, ::std::optional<int64_t> dim, ::std::optional<int64_t> mask_type);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mask, ::std::optional<int64_t> dim, ::std::optional<int64_t> mask_type);
+};
+
+struct TORCH_API _masked_softmax_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, ::std::optional<int64_t>, ::std::optional<int64_t>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_masked_softmax";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_masked_softmax.out(Tensor self, Tensor mask, int? dim=None, int? mask_type=None, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & mask, ::std::optional<int64_t> dim, ::std::optional<int64_t> mask_type, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mask, ::std::optional<int64_t> dim, ::std::optional<int64_t> mask_type, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mkldnn_reshape.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mkldnn_reshape.h
new file mode 100644
index 0000000000000000000000000000000000000000..179018685635e92a8774994c4318d74067f463c7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mkldnn_reshape.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_mkldnn_reshape_ops.h>
+
+namespace at {
+
+
+// aten::_mkldnn_reshape(Tensor self, int[] shape) -> Tensor
+inline at::Tensor _mkldnn_reshape(const at::Tensor & self, at::IntArrayRef shape) {
+    return at::_ops::_mkldnn_reshape::call(self, shape);
+}
+
+// aten::_mkldnn_reshape.out(Tensor self, int[] shape, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _mkldnn_reshape_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef shape) {
+    return at::_ops::_mkldnn_reshape_out::call(self, shape, out);
+}
+// aten::_mkldnn_reshape.out(Tensor self, int[] shape, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _mkldnn_reshape_outf(const at::Tensor & self, at::IntArrayRef shape, at::Tensor & out) {
+    return at::_ops::_mkldnn_reshape_out::call(self, shape, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mkldnn_transpose.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mkldnn_transpose.h
new file mode 100644
index 0000000000000000000000000000000000000000..99b2ba51d47151e4432478f79f22cd0b819fc343
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mkldnn_transpose.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_mkldnn_transpose_ops.h>
+
+namespace at {
+
+
+// aten::_mkldnn_transpose(Tensor self, int dim0, int dim1) -> Tensor
+inline at::Tensor _mkldnn_transpose(const at::Tensor & self, int64_t dim0, int64_t dim1) {
+    return at::_ops::_mkldnn_transpose::call(self, dim0, dim1);
+}
+
+// aten::_mkldnn_transpose_(Tensor(a!) self, int dim0, int dim1) -> Tensor(a!)
+inline at::Tensor & _mkldnn_transpose_(at::Tensor & self, int64_t dim0, int64_t dim1) {
+    return at::_ops::_mkldnn_transpose_::call(self, dim0, dim1);
+}
+
+// aten::_mkldnn_transpose.out(Tensor self, int dim0, int dim1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _mkldnn_transpose_out(at::Tensor & out, const at::Tensor & self, int64_t dim0, int64_t dim1) {
+    return at::_ops::_mkldnn_transpose_out::call(self, dim0, dim1, out);
+}
+// aten::_mkldnn_transpose.out(Tensor self, int dim0, int dim1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _mkldnn_transpose_outf(const at::Tensor & self, int64_t dim0, int64_t dim1, at::Tensor & out) {
+    return at::_ops::_mkldnn_transpose_out::call(self, dim0, dim1, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mps_convolution_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mps_convolution_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b1b37e423a0fe93149ab22f83edb2c7f44b60e0a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mps_convolution_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _mps_convolution_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups);
+TORCH_API at::Tensor & _mps_convolution_outf(const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, at::Tensor & out);
+TORCH_API at::Tensor & _mps_convolution_symint_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups);
+TORCH_API at::Tensor & _mps_convolution_symint_outf(const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mps_convolution_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mps_convolution_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..c8186b0bf45489788a307a7eb00f998b68f1157f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mps_convolution_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _mps_convolution_out_symint(const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mps_convolution_transpose.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mps_convolution_transpose.h
new file mode 100644
index 0000000000000000000000000000000000000000..41a15aa5c7727a84409e6368a48798fa7f151d75
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_mps_convolution_transpose.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_mps_convolution_transpose_ops.h>
+
+namespace at {
+
+
+// aten::_mps_convolution_transpose(Tensor self, Tensor weight, SymInt[] padding, SymInt[] output_padding, SymInt[] stride, SymInt[] dilation, SymInt groups) -> Tensor
+inline at::Tensor _mps_convolution_transpose(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef output_padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups) {
+    return at::_ops::_mps_convolution_transpose::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(output_padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor _mps_convolution_transpose(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef output_padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups) {
+    return at::_ops::_mps_convolution_transpose::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(output_padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups);
+  }
+}
+
+// aten::_mps_convolution_transpose(Tensor self, Tensor weight, SymInt[] padding, SymInt[] output_padding, SymInt[] stride, SymInt[] dilation, SymInt groups) -> Tensor
+inline at::Tensor _mps_convolution_transpose_symint(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups) {
+    return at::_ops::_mps_convolution_transpose::call(self, weight, padding, output_padding, stride, dilation, groups);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor _mps_convolution_transpose(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups) {
+    return at::_ops::_mps_convolution_transpose::call(self, weight, padding, output_padding, stride, dilation, groups);
+  }
+}
+
+// aten::_mps_convolution_transpose.out(Tensor self, Tensor weight, SymInt[] padding, SymInt[] output_padding, SymInt[] stride, SymInt[] dilation, SymInt groups, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _mps_convolution_transpose_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef output_padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups) {
+    return at::_ops::_mps_convolution_transpose_out::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(output_padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & _mps_convolution_transpose_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef output_padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups) {
+    return at::_ops::_mps_convolution_transpose_out::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(output_padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, out);
+  }
+}
+
+// aten::_mps_convolution_transpose.out(Tensor self, Tensor weight, SymInt[] padding, SymInt[] output_padding, SymInt[] stride, SymInt[] dilation, SymInt groups, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _mps_convolution_transpose_outf(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef output_padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, at::Tensor & out) {
+    return at::_ops::_mps_convolution_transpose_out::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(output_padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & _mps_convolution_transpose_outf(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef output_padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, at::Tensor & out) {
+    return at::_ops::_mps_convolution_transpose_out::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(output_padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, out);
+  }
+}
+
+// aten::_mps_convolution_transpose.out(Tensor self, Tensor weight, SymInt[] padding, SymInt[] output_padding, SymInt[] stride, SymInt[] dilation, SymInt groups, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _mps_convolution_transpose_symint_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups) {
+    return at::_ops::_mps_convolution_transpose_out::call(self, weight, padding, output_padding, stride, dilation, groups, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & _mps_convolution_transpose_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups) {
+    return at::_ops::_mps_convolution_transpose_out::call(self, weight, padding, output_padding, stride, dilation, groups, out);
+  }
+}
+
+// aten::_mps_convolution_transpose.out(Tensor self, Tensor weight, SymInt[] padding, SymInt[] output_padding, SymInt[] stride, SymInt[] dilation, SymInt groups, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _mps_convolution_transpose_symint_outf(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, at::Tensor & out) {
+    return at::_ops::_mps_convolution_transpose_out::call(self, weight, padding, output_padding, stride, dilation, groups, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & _mps_convolution_transpose_outf(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, at::Tensor & out) {
+    return at::_ops::_mps_convolution_transpose_out::call(self, weight, padding, output_padding, stride, dilation, groups, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_native_batch_norm_legit_no_training_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_native_batch_norm_legit_no_training_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..32ed3eea12574c983537f77ec4732f3fe621e3a6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_native_batch_norm_legit_no_training_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _batch_norm_legit_no_training(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, const at::Tensor & running_mean, const at::Tensor & running_var, double momentum, double eps);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _native_batch_norm_legit_no_training_out(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, const at::Tensor & running_mean, const at::Tensor & running_var, double momentum, double eps, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_native_multi_head_attention.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_native_multi_head_attention.h
new file mode 100644
index 0000000000000000000000000000000000000000..42c666dedb0150d8bf1de03969967354d2f102d9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_native_multi_head_attention.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_native_multi_head_attention_ops.h>
+
+namespace at {
+
+
+// aten::_native_multi_head_attention(Tensor query, Tensor key, Tensor value, int embed_dim, int num_head, Tensor qkv_weight, Tensor qkv_bias, Tensor proj_weight, Tensor proj_bias, Tensor? mask=None, bool need_weights=True, bool average_attn_weights=True, int? mask_type=None) -> (Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor> _native_multi_head_attention(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, int64_t embed_dim, int64_t num_head, const at::Tensor & qkv_weight, const at::Tensor & qkv_bias, const at::Tensor & proj_weight, const at::Tensor & proj_bias, const ::std::optional<at::Tensor> & mask={}, bool need_weights=true, bool average_attn_weights=true, ::std::optional<int64_t> mask_type=::std::nullopt) {
+    return at::_ops::_native_multi_head_attention::call(query, key, value, embed_dim, num_head, qkv_weight, qkv_bias, proj_weight, proj_bias, mask, need_weights, average_attn_weights, mask_type);
+}
+
+// aten::_native_multi_head_attention.out(Tensor query, Tensor key, Tensor value, int embed_dim, int num_head, Tensor qkv_weight, Tensor qkv_bias, Tensor proj_weight, Tensor proj_bias, Tensor? mask=None, bool need_weights=True, bool average_attn_weights=True, int? mask_type=None, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))
+inline ::std::tuple<at::Tensor &,at::Tensor &> _native_multi_head_attention_out(at::Tensor & out0, at::Tensor & out1, const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, int64_t embed_dim, int64_t num_head, const at::Tensor & qkv_weight, const at::Tensor & qkv_bias, const at::Tensor & proj_weight, const at::Tensor & proj_bias, const ::std::optional<at::Tensor> & mask={}, bool need_weights=true, bool average_attn_weights=true, ::std::optional<int64_t> mask_type=::std::nullopt) {
+    return at::_ops::_native_multi_head_attention_out::call(query, key, value, embed_dim, num_head, qkv_weight, qkv_bias, proj_weight, proj_bias, mask, need_weights, average_attn_weights, mask_type, out0, out1);
+}
+// aten::_native_multi_head_attention.out(Tensor query, Tensor key, Tensor value, int embed_dim, int num_head, Tensor qkv_weight, Tensor qkv_bias, Tensor proj_weight, Tensor proj_bias, Tensor? mask=None, bool need_weights=True, bool average_attn_weights=True, int? mask_type=None, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))
+inline ::std::tuple<at::Tensor &,at::Tensor &> _native_multi_head_attention_outf(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, int64_t embed_dim, int64_t num_head, const at::Tensor & qkv_weight, const at::Tensor & qkv_bias, const at::Tensor & proj_weight, const at::Tensor & proj_bias, const ::std::optional<at::Tensor> & mask, bool need_weights, bool average_attn_weights, ::std::optional<int64_t> mask_type, at::Tensor & out0, at::Tensor & out1) {
+    return at::_ops::_native_multi_head_attention_out::call(query, key, value, embed_dim, num_head, qkv_weight, qkv_bias, proj_weight, proj_bias, mask, need_weights, average_attn_weights, mask_type, out0, out1);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_native_multi_head_attention_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_native_multi_head_attention_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..1dfa0318cfc7c8001cd2b8abfa6d176229a3bd3c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_native_multi_head_attention_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> _native_multi_head_attention(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, int64_t embed_dim, int64_t num_head, const at::Tensor & qkv_weight, const at::Tensor & qkv_bias, const at::Tensor & proj_weight, const at::Tensor & proj_bias, const ::std::optional<at::Tensor> & mask={}, bool need_weights=true, bool average_attn_weights=true, ::std::optional<int64_t> mask_type=::std::nullopt);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_neg_view_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_neg_view_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..8d865ee32d39f92b446511ee9ecc81227911e393
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_neg_view_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _neg_view(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_jagged_dummy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_jagged_dummy.h
new file mode 100644
index 0000000000000000000000000000000000000000..6047509cc8d5f635acb7c4e4ca748f4fb8d54d9c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_jagged_dummy.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_nested_get_jagged_dummy_ops.h>
+
+namespace at {
+
+
+// aten::_nested_get_jagged_dummy(Tensor any) -> Tensor
+inline at::Tensor _nested_get_jagged_dummy(const at::Tensor & any) {
+    return at::_ops::_nested_get_jagged_dummy::call(any);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_max_seqlen_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_max_seqlen_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..c3cf204672f0da0f82bf74e5283a351132dd287c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_max_seqlen_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _nested_get_max_seqlen {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_nested_get_max_seqlen";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_nested_get_max_seqlen(Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_offsets_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_offsets_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..5abb091ecc3991ff4e1a92ae53f604557291c158
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_offsets_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _nested_get_offsets {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_nested_get_offsets";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_nested_get_offsets(Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_values_copy_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_values_copy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..cfd31dde53c7349da62a2054f930e8db30d45e15
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_get_values_copy_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _nested_get_values_copy_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor _nested_get_values_copy(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_select_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_select_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..1e38bc55e96d592667d1a51397539f85a4d2fc92
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_select_backward_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _nested_select_backward_symint(const at::Tensor & grad_output, const at::Tensor & self, int64_t dim, c10::SymInt index);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_from_mask.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_from_mask.h
new file mode 100644
index 0000000000000000000000000000000000000000..f02af7d29f82643df1ce850045f223ae47c59df1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_from_mask.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_nested_tensor_from_mask_ops.h>
+
+namespace at {
+
+
+// aten::_nested_tensor_from_mask(Tensor t, Tensor mask, bool mask_check=True) -> Tensor
+inline at::Tensor _nested_tensor_from_mask(const at::Tensor & t, const at::Tensor & mask, bool mask_check=true) {
+    return at::_ops::_nested_tensor_from_mask::call(t, mask, mask_check);
+}
+
+// aten::_nested_tensor_from_mask.out(Tensor t, Tensor mask, bool mask_check=True, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _nested_tensor_from_mask_out(at::Tensor & out, const at::Tensor & t, const at::Tensor & mask, bool mask_check=true) {
+    return at::_ops::_nested_tensor_from_mask_out::call(t, mask, mask_check, out);
+}
+// aten::_nested_tensor_from_mask.out(Tensor t, Tensor mask, bool mask_check=True, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _nested_tensor_from_mask_outf(const at::Tensor & t, const at::Tensor & mask, bool mask_check, at::Tensor & out) {
+    return at::_ops::_nested_tensor_from_mask_out::call(t, mask, mask_check, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_from_mask_left_aligned_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_from_mask_left_aligned_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..293c03d886fdf51e7066ee164bd3e9c306fa1d90
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_from_mask_left_aligned_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API bool NestedTensor_nested_tensor_from_mask_left_aligned(const at::Tensor & t, const at::Tensor & mask);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_from_mask_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_from_mask_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..ad6172e95ae10ed780ebbde674aa0f9527fdd6ab
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_from_mask_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _nested_tensor_from_mask_out(const at::Tensor & t, const at::Tensor & mask, bool mask_check, at::Tensor & out);
+TORCH_API at::Tensor NestedTensor_nested_tensor_from_mask(const at::Tensor & t, const at::Tensor & mask, bool mask_check=true);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_size_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_size_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..7caa8c2a758821e52ee0951ddbf4abab42d6c949
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_size_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _nested_tensor_size_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & _nested_tensor_size_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_storage_offsets_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_storage_offsets_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..e06bb6935a48845f13792602318b2d00e1b2879a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_storage_offsets_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _nested_tensor_storage_offsets_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor _nested_tensor_storage_offsets(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_strides_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_strides_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f608d6995970a277ae28f29f73c8b3d421e6acbc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_tensor_strides_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _nested_tensor_strides_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & _nested_tensor_strides_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_view_from_buffer_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_view_from_buffer_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..5c833088b9dcf591161445ca971587233f51302c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nested_view_from_buffer_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _nested_view_from_buffer {
+  using schema = at::Tensor (const 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 const char* name = "aten::_nested_view_from_buffer";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_nested_view_from_buffer(Tensor(a) self, Tensor nested_size, Tensor nested_strides, Tensor offsets) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & nested_size, const at::Tensor & nested_strides, const at::Tensor & offsets);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & nested_size, const at::Tensor & nested_strides, const at::Tensor & offsets);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnpack_available.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnpack_available.h
new file mode 100644
index 0000000000000000000000000000000000000000..1126e2234e075117da735b94e4d0783f3586cd89
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnpack_available.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_nnpack_available_ops.h>
+
+namespace at {
+
+
+// aten::_nnpack_available() -> bool
+inline bool _nnpack_available() {
+    return at::_ops::_nnpack_available::call();
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnpack_spatial_convolution.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnpack_spatial_convolution.h
new file mode 100644
index 0000000000000000000000000000000000000000..372bfb2602ed443849dc30a62f6ac13eabff4da8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnpack_spatial_convolution.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_nnpack_spatial_convolution_ops.h>
+
+namespace at {
+
+
+// aten::_nnpack_spatial_convolution(Tensor input, Tensor weight, Tensor? bias, SymInt[2] padding, SymInt[2] stride=1) -> Tensor
+inline at::Tensor _nnpack_spatial_convolution(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef padding, at::IntArrayRef stride=1) {
+    return at::_ops::_nnpack_spatial_convolution::call(input, weight, bias, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride));
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor _nnpack_spatial_convolution(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef padding, at::IntArrayRef stride=1) {
+    return at::_ops::_nnpack_spatial_convolution::call(input, weight, bias, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride));
+  }
+}
+
+// aten::_nnpack_spatial_convolution(Tensor input, Tensor weight, Tensor? bias, SymInt[2] padding, SymInt[2] stride=1) -> Tensor
+inline at::Tensor _nnpack_spatial_convolution_symint(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride=c10::SymInt(1)) {
+    return at::_ops::_nnpack_spatial_convolution::call(input, weight, bias, padding, stride);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor _nnpack_spatial_convolution(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride=c10::SymInt(1)) {
+    return at::_ops::_nnpack_spatial_convolution::call(input, weight, bias, padding, stride);
+  }
+}
+
+// aten::_nnpack_spatial_convolution.out(Tensor input, Tensor weight, Tensor? bias, SymInt[2] padding, SymInt[2] stride=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _nnpack_spatial_convolution_out(at::Tensor & out, const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef padding, at::IntArrayRef stride=1) {
+    return at::_ops::_nnpack_spatial_convolution_out::call(input, weight, bias, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & _nnpack_spatial_convolution_out(at::Tensor & out, const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef padding, at::IntArrayRef stride=1) {
+    return at::_ops::_nnpack_spatial_convolution_out::call(input, weight, bias, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), out);
+  }
+}
+
+// aten::_nnpack_spatial_convolution.out(Tensor input, Tensor weight, Tensor? bias, SymInt[2] padding, SymInt[2] stride=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _nnpack_spatial_convolution_outf(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef padding, at::IntArrayRef stride, at::Tensor & out) {
+    return at::_ops::_nnpack_spatial_convolution_out::call(input, weight, bias, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & _nnpack_spatial_convolution_outf(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef padding, at::IntArrayRef stride, at::Tensor & out) {
+    return at::_ops::_nnpack_spatial_convolution_out::call(input, weight, bias, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), out);
+  }
+}
+
+// aten::_nnpack_spatial_convolution.out(Tensor input, Tensor weight, Tensor? bias, SymInt[2] padding, SymInt[2] stride=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _nnpack_spatial_convolution_symint_out(at::Tensor & out, const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride=c10::SymInt(1)) {
+    return at::_ops::_nnpack_spatial_convolution_out::call(input, weight, bias, padding, stride, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & _nnpack_spatial_convolution_out(at::Tensor & out, const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride=c10::SymInt(1)) {
+    return at::_ops::_nnpack_spatial_convolution_out::call(input, weight, bias, padding, stride, out);
+  }
+}
+
+// aten::_nnpack_spatial_convolution.out(Tensor input, Tensor weight, Tensor? bias, SymInt[2] padding, SymInt[2] stride=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _nnpack_spatial_convolution_symint_outf(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, at::Tensor & out) {
+    return at::_ops::_nnpack_spatial_convolution_out::call(input, weight, bias, padding, stride, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & _nnpack_spatial_convolution_outf(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, at::Tensor & out) {
+    return at::_ops::_nnpack_spatial_convolution_out::call(input, weight, bias, padding, stride, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnpack_spatial_convolution_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnpack_spatial_convolution_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..c7b1903263cad251ec626cd81a75097531fa5bf5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnpack_spatial_convolution_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _nnpack_spatial_convolution {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_nnpack_spatial_convolution";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_nnpack_spatial_convolution(Tensor input, Tensor weight, Tensor? bias, SymInt[2] padding, SymInt[2] stride=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride);
+};
+
+struct TORCH_API _nnpack_spatial_convolution_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_nnpack_spatial_convolution";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_nnpack_spatial_convolution.out(Tensor input, Tensor weight, Tensor? bias, SymInt[2] padding, SymInt[2] stride=1, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnz.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnz.h
new file mode 100644
index 0000000000000000000000000000000000000000..d105722242ac7b04cc3b2ca5e5fa32d0646b155e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_nnz.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_nnz_ops.h>
+
+namespace at {
+
+
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pack_padded_sequence.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pack_padded_sequence.h
new file mode 100644
index 0000000000000000000000000000000000000000..3826d198de9d60af8f0739d81ea87b10a1a738ac
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pack_padded_sequence.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_pack_padded_sequence_ops.h>
+
+namespace at {
+
+
+// aten::_pack_padded_sequence(Tensor input, Tensor lengths, bool batch_first) -> (Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor> _pack_padded_sequence(const at::Tensor & input, const at::Tensor & lengths, bool batch_first) {
+    return at::_ops::_pack_padded_sequence::call(input, lengths, batch_first);
+}
+
+// aten::_pack_padded_sequence.out(Tensor input, Tensor lengths, bool batch_first, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))
+inline ::std::tuple<at::Tensor &,at::Tensor &> _pack_padded_sequence_out(at::Tensor & out0, at::Tensor & out1, const at::Tensor & input, const at::Tensor & lengths, bool batch_first) {
+    return at::_ops::_pack_padded_sequence_out::call(input, lengths, batch_first, out0, out1);
+}
+// aten::_pack_padded_sequence.out(Tensor input, Tensor lengths, bool batch_first, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))
+inline ::std::tuple<at::Tensor &,at::Tensor &> _pack_padded_sequence_outf(const at::Tensor & input, const at::Tensor & lengths, bool batch_first, at::Tensor & out0, at::Tensor & out1) {
+    return at::_ops::_pack_padded_sequence_out::call(input, lengths, batch_first, out0, out1);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pad_circular_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pad_circular_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..67d143bd325af7fd481da5d16e3dd509d4cad56c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pad_circular_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _pad_circular(const at::Tensor & self, at::IntArrayRef pad);
+TORCH_API at::Tensor _pad_circular_symint(const at::Tensor & self, c10::SymIntArrayRef pad);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pad_enum_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pad_enum_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..72203b20c81d8542936bf11648d97e3c93d1bf25
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pad_enum_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _pad_enum(const at::Tensor & self, at::IntArrayRef pad, int64_t mode, ::std::optional<double> value=::std::nullopt);
+TORCH_API at::Tensor _pad_enum_symint(const at::Tensor & self, c10::SymIntArrayRef pad, int64_t mode, ::std::optional<double> value=::std::nullopt);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pad_packed_sequence_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pad_packed_sequence_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8ca2fb2f79efd12da548a9e9e55fb646ec7f9afc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pad_packed_sequence_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,at::Tensor> _pad_packed_sequence(const at::Tensor & data, const at::Tensor & batch_sizes, bool batch_first, const at::Scalar & padding_value, int64_t total_length);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pdist_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pdist_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e0c350afb3e712cc34b53108d8b3295911533b56
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_pdist_backward_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _pdist_backward(const at::Tensor & grad, const at::Tensor & self, double p, const at::Tensor & pdist);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_print_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_print_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e85728ec26480e5cd20ea8c052e4ec92cbe768d4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_print_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void _print(c10::string_view s);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_print_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_print_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..fd9626ad94379db12a4ed1ff306eaad2f744a730
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_print_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _print {
+  using schema = void (c10::string_view);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_print";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_print(str s) -> ()";
+  static void call(c10::string_view s);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, c10::string_view s);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_propagate_xla_data_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_propagate_xla_data_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..7bfc2a0952dc8837afceafa79f44632c8da395f5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_propagate_xla_data_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void _propagate_xla_data(const at::Tensor & input, const at::Tensor & output);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_propagate_xla_data_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_propagate_xla_data_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..4e0fba81562e6e1592792e7ab73eeec9105b3b00
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_propagate_xla_data_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _propagate_xla_data {
+  using schema = void (const at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_propagate_xla_data";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_propagate_xla_data(Tensor input, Tensor output) -> ()";
+  static void call(const at::Tensor & input, const at::Tensor & output);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & output);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias.h
new file mode 100644
index 0000000000000000000000000000000000000000..fc6cfee04d508a918dcc7f9f5fff6c37517ae51c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias.h
@@ -0,0 +1,53 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_reshape_alias_ops.h>
+
+namespace at {
+
+
+// aten::_reshape_alias(Tensor(a) self, SymInt[] size, SymInt[] stride) -> Tensor(a)
+inline at::Tensor _reshape_alias(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride) {
+    return at::_ops::_reshape_alias::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride));
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor _reshape_alias(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride) {
+    return at::_ops::_reshape_alias::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride));
+  }
+}
+
+// aten::_reshape_alias(Tensor(a) self, SymInt[] size, SymInt[] stride) -> Tensor(a)
+inline at::Tensor _reshape_alias_symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride) {
+    return at::_ops::_reshape_alias::call(self, size, stride);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor _reshape_alias(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride) {
+    return at::_ops::_reshape_alias::call(self, size, stride);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias_copy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias_copy.h
new file mode 100644
index 0000000000000000000000000000000000000000..3b35f347756d285b03cfd09af43dce0fac688785
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias_copy.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_reshape_alias_copy_ops.h>
+
+namespace at {
+
+
+// aten::_reshape_alias_copy(Tensor self, SymInt[] size, SymInt[] stride) -> Tensor
+inline at::Tensor _reshape_alias_copy(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride) {
+    return at::_ops::_reshape_alias_copy::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride));
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor _reshape_alias_copy(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride) {
+    return at::_ops::_reshape_alias_copy::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride));
+  }
+}
+
+// aten::_reshape_alias_copy(Tensor self, SymInt[] size, SymInt[] stride) -> Tensor
+inline at::Tensor _reshape_alias_copy_symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride) {
+    return at::_ops::_reshape_alias_copy::call(self, size, stride);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor _reshape_alias_copy(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride) {
+    return at::_ops::_reshape_alias_copy::call(self, size, stride);
+  }
+}
+
+// aten::_reshape_alias_copy.out(Tensor self, SymInt[] size, SymInt[] stride, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _reshape_alias_copy_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride) {
+    return at::_ops::_reshape_alias_copy_out::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & _reshape_alias_copy_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride) {
+    return at::_ops::_reshape_alias_copy_out::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride), out);
+  }
+}
+
+// aten::_reshape_alias_copy.out(Tensor self, SymInt[] size, SymInt[] stride, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _reshape_alias_copy_outf(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride, at::Tensor & out) {
+    return at::_ops::_reshape_alias_copy_out::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & _reshape_alias_copy_outf(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride, at::Tensor & out) {
+    return at::_ops::_reshape_alias_copy_out::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride), out);
+  }
+}
+
+// aten::_reshape_alias_copy.out(Tensor self, SymInt[] size, SymInt[] stride, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _reshape_alias_copy_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride) {
+    return at::_ops::_reshape_alias_copy_out::call(self, size, stride, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & _reshape_alias_copy_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride) {
+    return at::_ops::_reshape_alias_copy_out::call(self, size, stride, out);
+  }
+}
+
+// aten::_reshape_alias_copy.out(Tensor self, SymInt[] size, SymInt[] stride, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _reshape_alias_copy_symint_outf(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, at::Tensor & out) {
+    return at::_ops::_reshape_alias_copy_out::call(self, size, stride, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & _reshape_alias_copy_outf(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, at::Tensor & out) {
+    return at::_ops::_reshape_alias_copy_out::call(self, size, stride, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..584c3b1f737bcf6678379f4395bde283f0ef3f2c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias_cuda_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _reshape_alias(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride);
+TORCH_API at::Tensor _reshape_alias_symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..00ac62d6ed2dbe916b996f7ab025bf2dd6ce96d6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_reshape_alias_meta_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _reshape_alias(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride);
+TORCH_API at::Tensor _reshape_alias_symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_rowwise_prune.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_rowwise_prune.h
new file mode 100644
index 0000000000000000000000000000000000000000..8fb8761c68007b8c99bf40c4cccadbc4798761e2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_rowwise_prune.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_rowwise_prune_ops.h>
+
+namespace at {
+
+
+// aten::_rowwise_prune(Tensor weight, Tensor mask, ScalarType compressed_indices_dtype) -> (Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor> _rowwise_prune(const at::Tensor & weight, const at::Tensor & mask, at::ScalarType compressed_indices_dtype) {
+    return at::_ops::_rowwise_prune::call(weight, mask, compressed_indices_dtype);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_attention_math_for_mps_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_attention_math_for_mps_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..0463f07d0019e4a3a7c2e520af1d14cb76e15912
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_attention_math_for_mps_native.h
@@ -0,0 +1,25 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 {
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_efficient_attention.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_efficient_attention.h
new file mode 100644
index 0000000000000000000000000000000000000000..b1a7d306565d2ff7b53c4c3c7c2ae10b21ed2871
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_efficient_attention.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_scaled_dot_product_efficient_attention_ops.h>
+
+namespace at {
+
+
+// aten::_scaled_dot_product_efficient_attention(Tensor query, Tensor key, Tensor value, Tensor? attn_bias, bool compute_log_sumexp, float dropout_p=0.0, bool is_causal=False, *, float? scale=None) -> (Tensor output, Tensor log_sumexp, Tensor philox_seed, Tensor philox_offset)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_efficient_attention(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const ::std::optional<at::Tensor> & attn_bias, bool compute_log_sumexp, double dropout_p=0.0, bool is_causal=false, ::std::optional<double> scale=::std::nullopt) {
+    return at::_ops::_scaled_dot_product_efficient_attention::call(query, key, value, attn_bias, compute_log_sumexp, dropout_p, is_causal, scale);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_efficient_attention_backward.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_efficient_attention_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..1ea566402db050c99923cf3eeb664d21329c68da
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_efficient_attention_backward.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_scaled_dot_product_efficient_attention_backward_ops.h>
+
+namespace at {
+
+
+// aten::_scaled_dot_product_efficient_attention_backward(Tensor grad_out_, Tensor query, Tensor key, Tensor value, Tensor attn_bias, Tensor out, Tensor logsumexp, Tensor philox_seed, Tensor philox_offset, float dropout_p, bool[4] grad_input_mask, bool is_causal=False, *, float? scale=None) -> (Tensor, Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_efficient_attention_backward(const at::Tensor & grad_out_, const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const at::Tensor & attn_bias, const at::Tensor & out, const at::Tensor & logsumexp, const at::Tensor & philox_seed, const at::Tensor & philox_offset, double dropout_p, ::std::array<bool,4> grad_input_mask, bool is_causal=false, ::std::optional<double> scale=::std::nullopt) {
+    return at::_ops::_scaled_dot_product_efficient_attention_backward::call(grad_out_, query, key, value, attn_bias, out, logsumexp, philox_seed, philox_offset, dropout_p, grad_input_mask, is_causal, scale);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_flash_attention_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_flash_attention_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..d741f786f18d618ac6bbfc5984704a8c5ea946a0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_flash_attention_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,c10::SymInt,c10::SymInt,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_flash_attention_cuda(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, double dropout_p=0.0, bool is_causal=false, bool return_debug_mask=false, ::std::optional<double> scale=::std::nullopt);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,c10::SymInt,c10::SymInt,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_flash_attention_nestedtensor_cuda(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, double dropout_p=0.0, bool is_causal=false, bool return_debug_mask=false, ::std::optional<double> scale=::std::nullopt);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,c10::SymInt,c10::SymInt,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_flash_attention_cuda_quantized(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const ::std::optional<at::Tensor> & q_descale, const ::std::optional<at::Tensor> & k_descale, const ::std::optional<at::Tensor> & v_descale, double dropout_p=0.0, bool is_causal=false, bool return_debug_mask=false, ::std::optional<double> scale=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_fused_attention_overrideable_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_fused_attention_overrideable_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..2fe1a4c7aecef2ddb3655a34a879343aeb342956
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_dot_product_fused_attention_overrideable_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,c10::SymInt,c10::SymInt,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_fused_attention_overrideable(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const ::std::optional<at::Tensor> & attn_bias={}, double dropout_p=0.0, bool is_causal=false, bool return_debug_mask=false, ::std::optional<double> scale=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_mm_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_mm_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..cbc1d67cf5cd8b2b665a9fa16ea1bdf1a7b300b4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_mm_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _scaled_mm {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, ::std::optional<at::ScalarType>, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_scaled_mm";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_scaled_mm(Tensor self, Tensor mat2, Tensor scale_a, Tensor scale_b, Tensor? bias=None, Tensor? scale_result=None, ScalarType? out_dtype=None, bool use_fast_accum=False) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & mat2, const at::Tensor & scale_a, const at::Tensor & scale_b, const ::std::optional<at::Tensor> & bias, const ::std::optional<at::Tensor> & scale_result, ::std::optional<at::ScalarType> out_dtype, bool use_fast_accum);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat2, const at::Tensor & scale_a, const at::Tensor & scale_b, const ::std::optional<at::Tensor> & bias, const ::std::optional<at::Tensor> & scale_result, ::std::optional<at::ScalarType> out_dtype, bool use_fast_accum);
+};
+
+struct TORCH_API _scaled_mm_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, ::std::optional<at::ScalarType>, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_scaled_mm";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_scaled_mm.out(Tensor self, Tensor mat2, Tensor scale_a, Tensor scale_b, Tensor? bias=None, Tensor? scale_result=None, ScalarType? out_dtype=None, bool use_fast_accum=False, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & mat2, const at::Tensor & scale_a, const at::Tensor & scale_b, const ::std::optional<at::Tensor> & bias, const ::std::optional<at::Tensor> & scale_result, ::std::optional<at::ScalarType> out_dtype, bool use_fast_accum, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat2, const at::Tensor & scale_a, const at::Tensor & scale_b, const ::std::optional<at::Tensor> & bias, const ::std::optional<at::Tensor> & scale_result, ::std::optional<at::ScalarType> out_dtype, bool use_fast_accum, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_mm_v2.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_mm_v2.h
new file mode 100644
index 0000000000000000000000000000000000000000..31f55f1c0e10ee0270d8a6039f2cf3baa746caee
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_mm_v2.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_scaled_mm_v2_ops.h>
+
+namespace at {
+
+
+// aten::_scaled_mm_v2(Tensor self, Tensor mat2, Tensor[] scale_a, int[] recipe_a, int[] swizzle_a, Tensor[] scale_b, int[] recipe_b, int[] swizzle_b, Tensor? bias, ScalarType? out_dtype, int[] contraction_dim=[], bool use_fast_accum=False) -> Tensor
+inline at::Tensor _scaled_mm_v2(const at::Tensor & self, const at::Tensor & mat2, at::TensorList scale_a, at::IntArrayRef recipe_a, at::IntArrayRef swizzle_a, at::TensorList scale_b, at::IntArrayRef recipe_b, at::IntArrayRef swizzle_b, const ::std::optional<at::Tensor> & bias, ::std::optional<at::ScalarType> out_dtype, at::IntArrayRef contraction_dim={}, bool use_fast_accum=false) {
+    return at::_ops::_scaled_mm_v2::call(self, mat2, scale_a, recipe_a, swizzle_a, scale_b, recipe_b, swizzle_b, bias, out_dtype, contraction_dim, use_fast_accum);
+}
+
+// aten::_scaled_mm_v2.out(Tensor self, Tensor mat2, Tensor[] scale_a, int[] recipe_a, int[] swizzle_a, Tensor[] scale_b, int[] recipe_b, int[] swizzle_b, Tensor? bias, ScalarType? out_dtype, int[] contraction_dim=[], bool use_fast_accum=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _scaled_mm_v2_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat2, at::TensorList scale_a, at::IntArrayRef recipe_a, at::IntArrayRef swizzle_a, at::TensorList scale_b, at::IntArrayRef recipe_b, at::IntArrayRef swizzle_b, const ::std::optional<at::Tensor> & bias, ::std::optional<at::ScalarType> out_dtype, at::IntArrayRef contraction_dim={}, bool use_fast_accum=false) {
+    return at::_ops::_scaled_mm_v2_out::call(self, mat2, scale_a, recipe_a, swizzle_a, scale_b, recipe_b, swizzle_b, bias, out_dtype, contraction_dim, use_fast_accum, out);
+}
+// aten::_scaled_mm_v2.out(Tensor self, Tensor mat2, Tensor[] scale_a, int[] recipe_a, int[] swizzle_a, Tensor[] scale_b, int[] recipe_b, int[] swizzle_b, Tensor? bias, ScalarType? out_dtype, int[] contraction_dim=[], bool use_fast_accum=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _scaled_mm_v2_outf(const at::Tensor & self, const at::Tensor & mat2, at::TensorList scale_a, at::IntArrayRef recipe_a, at::IntArrayRef swizzle_a, at::TensorList scale_b, at::IntArrayRef recipe_b, at::IntArrayRef swizzle_b, const ::std::optional<at::Tensor> & bias, ::std::optional<at::ScalarType> out_dtype, at::IntArrayRef contraction_dim, bool use_fast_accum, at::Tensor & out) {
+    return at::_ops::_scaled_mm_v2_out::call(self, mat2, scale_a, recipe_a, swizzle_a, scale_b, recipe_b, swizzle_b, bias, out_dtype, contraction_dim, use_fast_accum, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_mm_v2_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_mm_v2_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9f924b7455cf5c3d188c0d0ccabf6daca3d1ddd1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_scaled_mm_v2_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _scaled_mm_v2(const at::Tensor & self, const at::Tensor & mat2, at::TensorList scale_a, at::IntArrayRef recipe_a, at::IntArrayRef swizzle_a, at::TensorList scale_b, at::IntArrayRef recipe_b, at::IntArrayRef swizzle_b, const ::std::optional<at::Tensor> & bias, ::std::optional<at::ScalarType> out_dtype, at::IntArrayRef contraction_dim={}, bool use_fast_accum=false);
+TORCH_API at::Tensor & _scaled_mm_v2_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat2, at::TensorList scale_a, at::IntArrayRef recipe_a, at::IntArrayRef swizzle_a, at::TensorList scale_b, at::IntArrayRef recipe_b, at::IntArrayRef swizzle_b, const ::std::optional<at::Tensor> & bias, ::std::optional<at::ScalarType> out_dtype, at::IntArrayRef contraction_dim={}, bool use_fast_accum=false);
+TORCH_API at::Tensor & _scaled_mm_v2_outf(const at::Tensor & self, const at::Tensor & mat2, at::TensorList scale_a, at::IntArrayRef recipe_a, at::IntArrayRef swizzle_a, at::TensorList scale_b, at::IntArrayRef recipe_b, at::IntArrayRef swizzle_b, const ::std::optional<at::Tensor> & bias, ::std::optional<at::ScalarType> out_dtype, at::IntArrayRef contraction_dim, bool use_fast_accum, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sobol_engine_draw.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sobol_engine_draw.h
new file mode 100644
index 0000000000000000000000000000000000000000..e4d8e017ed522b17a79a8cb80e2b77e9d3c95d31
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sobol_engine_draw.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_sobol_engine_draw_ops.h>
+
+namespace at {
+
+
+// aten::_sobol_engine_draw(Tensor quasi, int n, Tensor sobolstate, int dimension, int num_generated, ScalarType? dtype) -> (Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor> _sobol_engine_draw(const at::Tensor & quasi, int64_t n, const at::Tensor & sobolstate, int64_t dimension, int64_t num_generated, ::std::optional<at::ScalarType> dtype) {
+    return at::_ops::_sobol_engine_draw::call(quasi, n, sobolstate, dimension, num_generated, dtype);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sobol_engine_ff_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sobol_engine_ff_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b5c2467e4bb6cefef916b6b7fa542b19381812c3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sobol_engine_ff_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor & _sobol_engine_ff_(at::Tensor & self, int64_t n, const at::Tensor & sobolstate, int64_t dimension, int64_t num_generated);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_softmax_backward_data.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_softmax_backward_data.h
new file mode 100644
index 0000000000000000000000000000000000000000..2f135c72946863edc53594cfe1497919b26277c1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_softmax_backward_data.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_softmax_backward_data_ops.h>
+
+namespace at {
+
+
+// aten::_softmax_backward_data(Tensor grad_output, Tensor output, int dim, ScalarType input_dtype) -> Tensor
+inline at::Tensor _softmax_backward_data(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype) {
+    return at::_ops::_softmax_backward_data::call(grad_output, output, dim, input_dtype);
+}
+
+// aten::_softmax_backward_data.out(Tensor grad_output, Tensor output, int dim, ScalarType input_dtype, *, Tensor(a!) grad_input) -> Tensor(a!)
+inline at::Tensor & _softmax_backward_data_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype) {
+    return at::_ops::_softmax_backward_data_out::call(grad_output, output, dim, input_dtype, grad_input);
+}
+// aten::_softmax_backward_data.out(Tensor grad_output, Tensor output, int dim, ScalarType input_dtype, *, Tensor(a!) grad_input) -> Tensor(a!)
+inline at::Tensor & _softmax_backward_data_outf(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype, at::Tensor & grad_input) {
+    return at::_ops::_softmax_backward_data_out::call(grad_output, output, dim, input_dtype, grad_input);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_softmax_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_softmax_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..2823bf50d418cd125d2ae4a38df43b5b981e1c08
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_softmax_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _softmax(const at::Tensor & self, int64_t dim, bool half_to_float);
+TORCH_API at::Tensor & _softmax_out(at::Tensor & out, const at::Tensor & self, int64_t dim, bool half_to_float);
+TORCH_API at::Tensor & _softmax_outf(const at::Tensor & self, int64_t dim, bool half_to_float, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_softmax_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_softmax_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..05f132037ed7b489bced4f50558440a0de2fc42d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_softmax_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _softmax(const at::Tensor & self, int64_t dim, bool half_to_float);
+TORCH_API at::Tensor & _softmax_out(at::Tensor & out, const at::Tensor & self, int64_t dim, bool half_to_float);
+TORCH_API at::Tensor & _softmax_outf(const at::Tensor & self, int64_t dim, bool half_to_float, at::Tensor & out);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_broadcast_to_copy_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_broadcast_to_copy_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e9d198d9191643e221ec9a58a8e88d36c2879ccf
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_broadcast_to_copy_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & _sparse_broadcast_to_copy_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef size);
+TORCH_API at::Tensor & _sparse_broadcast_to_copy_outf(const at::Tensor & self, at::IntArrayRef size, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_bsc_tensor_unsafe_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_bsc_tensor_unsafe_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..475b1e512be42f5042d393172c9cf21351cc984b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_bsc_tensor_unsafe_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _sparse_bsc_tensor_unsafe {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, at::IntArrayRef, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_bsc_tensor_unsafe";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_sparse_bsc_tensor_unsafe(Tensor ccol_indices, Tensor row_indices, Tensor values, int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & ccol_indices, const at::Tensor & row_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & ccol_indices, const at::Tensor & row_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_bsr_tensor_unsafe_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_bsr_tensor_unsafe_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..feaeefa3f7988f82f70591d0f4542525a02c9ccd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_bsr_tensor_unsafe_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor _sparse_bsr_tensor_unsafe(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, at::TensorOptions options={});
+TORCH_API at::Tensor _sparse_bsr_tensor_unsafe(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_unsafe_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_unsafe_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c8904318ac09d06dfc2675dcfbb4cf0110902f4d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_unsafe_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor _sparse_compressed_tensor_unsafe(const at::Tensor & compressed_indices, const at::Tensor & plain_indices, const at::Tensor & values, at::IntArrayRef size, at::TensorOptions options={});
+TORCH_API at::Tensor _sparse_compressed_tensor_unsafe(const at::Tensor & compressed_indices, const at::Tensor & plain_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+TORCH_API at::Tensor _sparse_compressed_tensor_unsafe_symint(const at::Tensor & compressed_indices, const at::Tensor & plain_indices, const at::Tensor & values, c10::SymIntArrayRef size, at::TensorOptions options={});
+TORCH_API at::Tensor _sparse_compressed_tensor_unsafe_symint(const at::Tensor & compressed_indices, const at::Tensor & plain_indices, const at::Tensor & values, c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_unsafe_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_unsafe_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..e97c11d3f3376bccc9770f874d598ce6b97a0bf0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_unsafe_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _sparse_compressed_tensor_unsafe_symint(const at::Tensor & compressed_indices, const at::Tensor & plain_indices, const at::Tensor & values, c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_unsafe_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_unsafe_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..d65d62c71eb7e8d854ba2bae1bd3bf9ba5b2f2a1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_unsafe_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _sparse_compressed_tensor_unsafe {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_compressed_tensor_unsafe";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_sparse_compressed_tensor_unsafe(Tensor compressed_indices, Tensor plain_indices, Tensor values, SymInt[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & compressed_indices, const at::Tensor & plain_indices, const at::Tensor & values, c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & compressed_indices, const at::Tensor & plain_indices, const at::Tensor & values, c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_with_dims_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_with_dims_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..cc94f16c0fd506dcfe3bd1044d4af10454ed4941
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_compressed_tensor_with_dims_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sparse_compressed_tensor_with_dims(int64_t nnz, int64_t dense_dim, at::IntArrayRef size, at::IntArrayRef blocksize, at::ScalarType index_dtype, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_unsafe_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_unsafe_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..5941b991a4230ea5dae0724423cc8c016201f17c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_unsafe_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _sparse_coo_tensor_unsafe {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_coo_tensor_unsafe";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_sparse_coo_tensor_unsafe(Tensor indices, Tensor values, SymInt[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, bool? is_coalesced=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & indices, const at::Tensor & values, c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<bool> is_coalesced);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & indices, const at::Tensor & values, c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<bool> is_coalesced);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_with_dims.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_with_dims.h
new file mode 100644
index 0000000000000000000000000000000000000000..437eff7c2eab1cc589e44b46afbaa767a31600e3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_with_dims.h
@@ -0,0 +1,49 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_sparse_coo_tensor_with_dims_ops.h>
+
+namespace at {
+
+
+// aten::_sparse_coo_tensor_with_dims(int sparse_dim, int dense_dim, int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=False) -> Tensor
+inline at::Tensor _sparse_coo_tensor_with_dims(int64_t sparse_dim, int64_t dense_dim, at::IntArrayRef size, at::TensorOptions options) {
+    return at::_ops::_sparse_coo_tensor_with_dims::call(sparse_dim, dense_dim, size, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+// aten::_sparse_coo_tensor_with_dims(int sparse_dim, int dense_dim, int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=False) -> Tensor
+inline at::Tensor _sparse_coo_tensor_with_dims(int64_t sparse_dim, int64_t dense_dim, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::_sparse_coo_tensor_with_dims::call(sparse_dim, dense_dim, size, dtype, layout, device, pin_memory);
+}
+
+// aten::_sparse_coo_tensor_with_dims.out(int sparse_dim, int dense_dim, int[] size, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _sparse_coo_tensor_with_dims_out(at::Tensor & out, int64_t sparse_dim, int64_t dense_dim, at::IntArrayRef size) {
+    return at::_ops::_sparse_coo_tensor_with_dims_out::call(sparse_dim, dense_dim, size, out);
+}
+// aten::_sparse_coo_tensor_with_dims.out(int sparse_dim, int dense_dim, int[] size, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _sparse_coo_tensor_with_dims_outf(int64_t sparse_dim, int64_t dense_dim, at::IntArrayRef size, at::Tensor & out) {
+    return at::_ops::_sparse_coo_tensor_with_dims_out::call(sparse_dim, dense_dim, size, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_with_dims_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_with_dims_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..44012b176eae1299d0e6052d6b3d75b2782edcc5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_with_dims_meta_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _sparse_coo_tensor_with_dims(int64_t sparse_dim, int64_t dense_dim, at::IntArrayRef size, at::TensorOptions options);
+TORCH_API at::Tensor _sparse_coo_tensor_with_dims(int64_t sparse_dim, int64_t dense_dim, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_with_dims_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_with_dims_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..03cad1d1d3f15ee3015d241f07062f9685233970
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_coo_tensor_with_dims_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _sparse_coo_tensor_with_dims {
+  using schema = at::Tensor (int64_t, int64_t, at::IntArrayRef, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_coo_tensor_with_dims";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_sparse_coo_tensor_with_dims(int sparse_dim, int dense_dim, int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=False) -> Tensor";
+  static at::Tensor call(int64_t sparse_dim, int64_t dense_dim, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, int64_t sparse_dim, int64_t dense_dim, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+struct TORCH_API _sparse_coo_tensor_with_dims_out {
+  using schema = at::Tensor & (int64_t, int64_t, at::IntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_coo_tensor_with_dims";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_sparse_coo_tensor_with_dims.out(int sparse_dim, int dense_dim, int[] size, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(int64_t sparse_dim, int64_t dense_dim, at::IntArrayRef size, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, int64_t sparse_dim, int64_t dense_dim, at::IntArrayRef size, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_csr_sum_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_csr_sum_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..35fab9ada5005d32546f76be5d771481b958d26a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_csr_sum_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & _sparse_csr_sum_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef dim, bool keepdim=false, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & _sparse_csr_sum_outf(const at::Tensor & self, at::IntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_csr_tensor_unsafe_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_csr_tensor_unsafe_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..ab77f13c1a5175f0b490c1b76f7679a43fb74762
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_csr_tensor_unsafe_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _sparse_csr_tensor_unsafe {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, at::IntArrayRef, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_csr_tensor_unsafe";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_sparse_csr_tensor_unsafe(Tensor crow_indices, Tensor col_indices, Tensor values, int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_backward_data_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_backward_data_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..988d130fcbe0daf6580d0b2395f5e91a3172b5b8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_backward_data_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & _sparse_log_softmax_backward_data_out(at::Tensor & out, const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, const at::Tensor & self);
+TORCH_API at::Tensor & _sparse_log_softmax_backward_data_outf(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, const at::Tensor & self, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3892290da4f50b30a6c7e6d7e47b5fe2a052e794
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & _sparse_log_softmax_out(at::Tensor & out, const at::Tensor & self, int64_t dim, bool half_to_float);
+TORCH_API at::Tensor & _sparse_log_softmax_outf(const at::Tensor & self, int64_t dim, bool half_to_float, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..a024d73dc676d7e32548e3bba054e390852f3eca
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_native.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _sparse_log_softmax(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor _sparse_log_softmax(const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & _sparse_log_softmax_out(const at::Tensor & self, int64_t dim, bool half_to_float, at::Tensor & out);
+TORCH_API at::Tensor log_softmax_sparse_cpu(const at::Tensor & self, int64_t dim, bool half_to_float);
+TORCH_API at::Tensor log_softmax_sparse_cuda(const at::Tensor & self, int64_t dim, bool half_to_float);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..9d5db02db2381c6ad24205ff14e29c66be4c1c3f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_log_softmax_ops.h
@@ -0,0 +1,67 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _sparse_log_softmax_int {
+  using schema = at::Tensor (const at::Tensor &, int64_t, ::std::optional<at::ScalarType>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_log_softmax";
+  static constexpr const char* overload_name = "int";
+  static constexpr const char* schema_str = "_sparse_log_softmax.int(Tensor self, int dim, ScalarType? dtype=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype);
+};
+
+struct TORCH_API _sparse_log_softmax_Dimname {
+  using schema = at::Tensor (const at::Tensor &, at::Dimname, ::std::optional<at::ScalarType>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_log_softmax";
+  static constexpr const char* overload_name = "Dimname";
+  static constexpr const char* schema_str = "_sparse_log_softmax.Dimname(Tensor self, Dimname dim, *, ScalarType? dtype=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype);
+};
+
+struct TORCH_API _sparse_log_softmax {
+  using schema = at::Tensor (const at::Tensor &, int64_t, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_log_softmax";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_sparse_log_softmax(Tensor self, int dim, bool half_to_float) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t dim, bool half_to_float);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, bool half_to_float);
+};
+
+struct TORCH_API _sparse_log_softmax_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_log_softmax";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_sparse_log_softmax.out(Tensor self, int dim, bool half_to_float, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, int64_t dim, bool half_to_float, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, bool half_to_float, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_mm.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_mm.h
new file mode 100644
index 0000000000000000000000000000000000000000..4670c3232ea2ca0d3b1b850ed2430f03d7f87698
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_mm.h
@@ -0,0 +1,41 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_sparse_mm_ops.h>
+
+namespace at {
+
+
+// aten::_sparse_mm(Tensor sparse, Tensor dense) -> Tensor
+inline at::Tensor _sparse_mm(const at::Tensor & sparse, const at::Tensor & dense) {
+    return at::_ops::_sparse_mm::call(sparse, dense);
+}
+
+// aten::_sparse_mm.reduce(Tensor sparse, Tensor dense, str reduce) -> Tensor
+inline at::Tensor _sparse_mm(const at::Tensor & sparse, const at::Tensor & dense, c10::string_view reduce) {
+    return at::_ops::_sparse_mm_reduce::call(sparse, dense, reduce);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_mm_reduce_impl_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_mm_reduce_impl_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..6580e5b91b36f793bbea0989c8af7aa214bca36e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_mm_reduce_impl_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> _sparse_mm_reduce_impl_sparse_csr_cpu(const at::Tensor & self, const at::Tensor & other, c10::string_view reduce);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_semi_structured_addmm_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_semi_structured_addmm_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8ba9054ff7481c1502e229ec74d57e621450ffe7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_semi_structured_addmm_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _sparse_semi_structured_addmm(const at::Tensor & input, const at::Tensor & mat1, const at::Tensor & mat1_meta, const at::Tensor & mat2, const at::Scalar & alpha=1, const at::Scalar & beta=1, ::std::optional<at::ScalarType> out_dtype=::std::nullopt);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_semi_structured_linear_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_semi_structured_linear_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..ca26cc6ec93e3c63cfa1997b36d38a77889fb66d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_semi_structured_linear_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _sparse_semi_structured_linear(const at::Tensor & input, const at::Tensor & weight, const at::Tensor & meta, const ::std::optional<at::Tensor> & bias={}, ::std::optional<c10::string_view> activation=::std::nullopt, ::std::optional<at::ScalarType> out_dtype=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_semi_structured_mm.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_semi_structured_mm.h
new file mode 100644
index 0000000000000000000000000000000000000000..b1830a4f19a4ec17a66a499c68e7becffb2b2c73
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_semi_structured_mm.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_sparse_semi_structured_mm_ops.h>
+
+namespace at {
+
+
+// aten::_sparse_semi_structured_mm(Tensor mat1, Tensor mat1_meta, Tensor mat2, *, ScalarType? out_dtype=None) -> Tensor
+inline at::Tensor _sparse_semi_structured_mm(const at::Tensor & mat1, const at::Tensor & mat1_meta, const at::Tensor & mat2, ::std::optional<at::ScalarType> out_dtype=::std::nullopt) {
+    return at::_ops::_sparse_semi_structured_mm::call(mat1, mat1_meta, mat2, out_dtype);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_softmax_backward_data_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_softmax_backward_data_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..1d108523f13a1ca95775e641a1f78be10f64d346
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_softmax_backward_data_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _sparse_softmax_backward_data {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, int64_t, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_softmax_backward_data";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_sparse_softmax_backward_data(Tensor grad_output, Tensor output, int dim, Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, const at::Tensor & self);
+};
+
+struct TORCH_API _sparse_softmax_backward_data_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, int64_t, const at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_sparse_softmax_backward_data";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_sparse_softmax_backward_data.out(Tensor grad_output, Tensor output, int dim, Tensor self, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, const at::Tensor & self, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, const at::Tensor & self, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_softmax_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_softmax_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ca3b615187aa872fbb029d07856cdf010228d06f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_softmax_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor _sparse_softmax(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor _sparse_softmax(const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_sparse_matmul_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_sparse_matmul_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a3c8b4017b4756f7c246447eb53d5c14792cdab6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_sparse_matmul_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & _sparse_sparse_matmul_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & _sparse_sparse_matmul_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_sum_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_sum_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..47443d0e40a2e51beef7af971e141c3ffe8be4fd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_sparse_sum_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor _sparse_sum(const at::Tensor & self, at::IntArrayRef dim);
+TORCH_API at::Tensor & _sparse_sum_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef dim);
+TORCH_API at::Tensor & _sparse_sum_outf(const at::Tensor & self, at::IntArrayRef dim, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_standard_gamma_grad_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_standard_gamma_grad_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ec7ea73c657a1f1f39f6776a5d394980b345f917
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_standard_gamma_grad_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _standard_gamma_grad(const at::Tensor & self, const at::Tensor & output);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_standard_gamma_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_standard_gamma_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..cf14e58abd2276d951e3b5fbe43844af838379ad
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_standard_gamma_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _standard_gamma_out(const at::Tensor & self, ::std::optional<at::Generator> generator, at::Tensor & out);
+TORCH_API at::Tensor _s_gamma_cpu(const at::Tensor & self, ::std::optional<at::Generator> generator=::std::nullopt);
+TORCH_API at::Tensor _s_gamma_cuda(const at::Tensor & self, ::std::optional<at::Generator> generator=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_standard_gamma_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_standard_gamma_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..dcf94a9fdc5b0277155543d7e0aa39f3e4b73289
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_standard_gamma_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _standard_gamma {
+  using schema = at::Tensor (const at::Tensor &, ::std::optional<at::Generator>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_standard_gamma";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_standard_gamma(Tensor self, Generator? generator=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, ::std::optional<at::Generator> generator);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<at::Generator> generator);
+};
+
+struct TORCH_API _standard_gamma_out {
+  using schema = at::Tensor & (const at::Tensor &, ::std::optional<at::Generator>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_standard_gamma";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_standard_gamma.out(Tensor self, Generator? generator=None, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, ::std::optional<at::Generator> generator, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<at::Generator> generator, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_ambiguous_defaults_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_ambiguous_defaults_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..19b9f5c64bd6a89d3a7ebe28878e42580b8c516d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_ambiguous_defaults_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_ambiguous_defaults(const at::Tensor & dummy, int64_t a=1, int64_t b=1);
+TORCH_API at::Tensor _test_ambiguous_defaults(const at::Tensor & dummy, int64_t a=2, c10::string_view b="2");
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_autograd_multiple_dispatch_view_copy_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_autograd_multiple_dispatch_view_copy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..0f49d9949148152f099a0e2578b9a35e882a6f98
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_autograd_multiple_dispatch_view_copy_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_autograd_multiple_dispatch_view_copy_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor _test_autograd_multiple_dispatch_view_copy(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_functorch_fallback_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_functorch_fallback_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..500c92f97bfd6fa83876e76085b7da24320bd68e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_functorch_fallback_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & _test_functorch_fallback_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & _test_functorch_fallback_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_functorch_fallback_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_functorch_fallback_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5de3a8f5915869e197dd329ebf39930cce96f283
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_functorch_fallback_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _test_functorch_fallback(const at::Tensor & self, const at::Tensor & other);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_optional_floatlist_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_optional_floatlist_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f0ecddf3e0510315944de0337a9442c789158245
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_optional_floatlist_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & _test_optional_floatlist_out(at::Tensor & out, const at::Tensor & values, ::std::optional<at::ArrayRef<double>> addends);
+TORCH_API at::Tensor & _test_optional_floatlist_outf(const at::Tensor & values, ::std::optional<at::ArrayRef<double>> addends, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_optional_intlist.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_optional_intlist.h
new file mode 100644
index 0000000000000000000000000000000000000000..076eb8ced1db6a42344740fa94211dd5fc0db374
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_optional_intlist.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_test_optional_intlist_ops.h>
+
+namespace at {
+
+
+// aten::_test_optional_intlist(Tensor values, int[]? addends) -> Tensor
+inline at::Tensor _test_optional_intlist(const at::Tensor & values, at::OptionalIntArrayRef addends) {
+    return at::_ops::_test_optional_intlist::call(values, addends);
+}
+
+// aten::_test_optional_intlist.out(Tensor values, int[]? addends, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _test_optional_intlist_out(at::Tensor & out, const at::Tensor & values, at::OptionalIntArrayRef addends) {
+    return at::_ops::_test_optional_intlist_out::call(values, addends, out);
+}
+// aten::_test_optional_intlist.out(Tensor values, int[]? addends, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _test_optional_intlist_outf(const at::Tensor & values, at::OptionalIntArrayRef addends, at::Tensor & out) {
+    return at::_ops::_test_optional_intlist_out::call(values, addends, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_optional_intlist_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_optional_intlist_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..39a4c53adf9cd74f8fb07286093a8662f5215465
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_optional_intlist_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _test_optional_intlist {
+  using schema = at::Tensor (const at::Tensor &, at::OptionalIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_test_optional_intlist";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_test_optional_intlist(Tensor values, int[]? addends) -> Tensor";
+  static at::Tensor call(const at::Tensor & values, at::OptionalIntArrayRef addends);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & values, at::OptionalIntArrayRef addends);
+};
+
+struct TORCH_API _test_optional_intlist_out {
+  using schema = at::Tensor & (const at::Tensor &, at::OptionalIntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_test_optional_intlist";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_test_optional_intlist.out(Tensor values, int[]? addends, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & values, at::OptionalIntArrayRef addends, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & values, at::OptionalIntArrayRef addends, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_serialization_subcmul.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_serialization_subcmul.h
new file mode 100644
index 0000000000000000000000000000000000000000..a40b46210c1965b1c0db832ef154c8805017e214
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_serialization_subcmul.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_test_serialization_subcmul_ops.h>
+
+namespace at {
+
+
+// aten::_test_serialization_subcmul(Tensor self, Tensor other, Scalar alpha=1) -> Tensor
+inline at::Tensor _test_serialization_subcmul(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1) {
+    return at::_ops::_test_serialization_subcmul::call(self, other, alpha);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_string_default.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_string_default.h
new file mode 100644
index 0000000000000000000000000000000000000000..5e18ac89a579db889fc5a966b88c172fe815b796
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_string_default.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_test_string_default_ops.h>
+
+namespace at {
+
+
+// aten::_test_string_default(Tensor dummy, str a="\"'\\", str b='"\'\\') -> Tensor
+inline at::Tensor _test_string_default(const at::Tensor & dummy, c10::string_view a="\"'\\", c10::string_view b="\"'\\") {
+    return at::_ops::_test_string_default::call(dummy, a, b);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_warn_in_autograd.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_warn_in_autograd.h
new file mode 100644
index 0000000000000000000000000000000000000000..1d8a6762c8e76dc57cefca8e894cf4b125156737
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_warn_in_autograd.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_test_warn_in_autograd_ops.h>
+
+namespace at {
+
+
+// aten::_test_warn_in_autograd(Tensor self) -> Tensor
+inline at::Tensor _test_warn_in_autograd(const at::Tensor & self) {
+    return at::_ops::_test_warn_in_autograd::call(self);
+}
+
+// aten::_test_warn_in_autograd.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _test_warn_in_autograd_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::_test_warn_in_autograd_out::call(self, out);
+}
+// aten::_test_warn_in_autograd.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _test_warn_in_autograd_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::_test_warn_in_autograd_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_warn_in_autograd_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_warn_in_autograd_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..ebc37dbc78804386bbea23d0abd0fe00d28939ab
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_test_warn_in_autograd_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _test_warn_in_autograd {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_test_warn_in_autograd";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_test_warn_in_autograd(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 _test_warn_in_autograd_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 const char* name = "aten::_test_warn_in_autograd";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_test_warn_in_autograd.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_differentiable_gru_cell_backward.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_differentiable_gru_cell_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..b75b5d612bc0780e4e04e8c4845469e8dd8ee1e5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_differentiable_gru_cell_backward.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_thnn_differentiable_gru_cell_backward_ops.h>
+
+namespace at {
+
+
+// aten::_thnn_differentiable_gru_cell_backward(Tensor grad_hy, Tensor input_gates, Tensor hidden_gates, Tensor hx, Tensor? input_bias, Tensor? hidden_bias) -> (Tensor, Tensor, Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,at::Tensor> _thnn_differentiable_gru_cell_backward(const at::Tensor & grad_hy, const at::Tensor & input_gates, const at::Tensor & hidden_gates, const at::Tensor & hx, const ::std::optional<at::Tensor> & input_bias, const ::std::optional<at::Tensor> & hidden_bias) {
+    return at::_ops::_thnn_differentiable_gru_cell_backward::call(grad_hy, input_gates, hidden_gates, hx, input_bias, hidden_bias);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_differentiable_gru_cell_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_differentiable_gru_cell_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..82c7c9509978411887397fa3011315ff8754870c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_differentiable_gru_cell_backward_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _thnn_differentiable_gru_cell_backward {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_thnn_differentiable_gru_cell_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_thnn_differentiable_gru_cell_backward(Tensor grad_hy, Tensor input_gates, Tensor hidden_gates, Tensor hx, Tensor? input_bias, Tensor? hidden_bias) -> (Tensor, Tensor, Tensor, Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,at::Tensor> call(const at::Tensor & grad_hy, const at::Tensor & input_gates, const at::Tensor & hidden_gates, const at::Tensor & hx, const ::std::optional<at::Tensor> & input_bias, const ::std::optional<at::Tensor> & hidden_bias);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_hy, const at::Tensor & input_gates, const at::Tensor & hidden_gates, const at::Tensor & hx, const ::std::optional<at::Tensor> & input_bias, const ::std::optional<at::Tensor> & hidden_bias);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_differentiable_lstm_cell_backward.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_differentiable_lstm_cell_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..37b93c0958a96acd8b599e9538edff655cb4f52a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_differentiable_lstm_cell_backward.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_thnn_differentiable_lstm_cell_backward_ops.h>
+
+namespace at {
+
+
+// aten::_thnn_differentiable_lstm_cell_backward(Tensor? grad_hy, Tensor? grad_cy, Tensor input_gates, Tensor hidden_gates, Tensor? input_bias, Tensor? hidden_bias, Tensor cx, Tensor cy) -> (Tensor, Tensor, Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,at::Tensor> _thnn_differentiable_lstm_cell_backward(const ::std::optional<at::Tensor> & grad_hy, const ::std::optional<at::Tensor> & grad_cy, const at::Tensor & input_gates, const at::Tensor & hidden_gates, const ::std::optional<at::Tensor> & input_bias, const ::std::optional<at::Tensor> & hidden_bias, const at::Tensor & cx, const at::Tensor & cy) {
+    return at::_ops::_thnn_differentiable_lstm_cell_backward::call(grad_hy, grad_cy, input_gates, hidden_gates, input_bias, hidden_bias, cx, cy);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_fused_gru_cell_backward_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_fused_gru_cell_backward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8e5a6b83a47a625746ef939cdf90adf6ce2720f0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_fused_gru_cell_backward_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,at::Tensor> _thnn_fused_gru_cell_backward(const at::Tensor & grad_hy, const at::Tensor & workspace, bool has_bias);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_fused_lstm_cell_backward_impl_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_fused_lstm_cell_backward_impl_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a36190638a0aedfa797c4214f926b00354915df4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_fused_lstm_cell_backward_impl_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _thnn_fused_lstm_cell_backward_impl_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const ::std::optional<at::Tensor> & grad_hy, const ::std::optional<at::Tensor> & grad_cy, const at::Tensor & cx, const at::Tensor & cy, const at::Tensor & workspace, bool has_bias);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _thnn_fused_lstm_cell_backward_impl_outf(const ::std::optional<at::Tensor> & grad_hy, const ::std::optional<at::Tensor> & grad_cy, const at::Tensor & cx, const at::Tensor & cy, const at::Tensor & workspace, bool has_bias, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_fused_lstm_cell_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_fused_lstm_cell_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b0cf63010e5c0e9bf871f4b816db8979f7a7398e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_thnn_fused_lstm_cell_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> _thnn_fused_lstm_cell(const at::Tensor & input_gates, const at::Tensor & hidden_gates, const at::Tensor & cx, const ::std::optional<at::Tensor> & input_bias={}, const ::std::optional<at::Tensor> & hidden_bias={});
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_dense_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_dense_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..594030297373acaba02f58f9b36641754d421c42
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_dense_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _to_dense_out(at::Tensor & out, const at::Tensor & self, ::std::optional<at::ScalarType> dtype=::std::nullopt, ::std::optional<bool> masked_grad=::std::nullopt);
+TORCH_API at::Tensor & _to_dense_outf(const at::Tensor & self, ::std::optional<at::ScalarType> dtype, ::std::optional<bool> masked_grad, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_bsc_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_bsc_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..87f7dfc1ff2ece7eaba0a4cbd320011db6b52c81
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_bsc_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _to_sparse_bsc_out(const at::Tensor & self, at::IntArrayRef blocksize, ::std::optional<int64_t> dense_dim, at::Tensor & out);
+TORCH_API at::Tensor dense_to_sparse_bsc(const at::Tensor & self, at::IntArrayRef blocksize, ::std::optional<int64_t> dense_dim=::std::nullopt);
+TORCH_API at::Tensor coo_to_sparse_bsc(const at::Tensor & self, at::IntArrayRef blocksize, ::std::optional<int64_t> dense_dim=::std::nullopt);
+TORCH_API at::Tensor sparse_compressed_to_sparse_bsc(const at::Tensor & self, at::IntArrayRef blocksize, ::std::optional<int64_t> dense_dim=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_bsr_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_bsr_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..cfb3cb03ff9a8114fd6e4c087e6915f18cbd93c8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_bsr_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor _to_sparse_bsr(const at::Tensor & self, at::IntArrayRef blocksize, ::std::optional<int64_t> dense_dim=::std::nullopt);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_csr_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_csr_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..7b1ff92406ac8fb7b7cf6ec034257120adacf3c6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_csr_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _to_sparse_csr_out(at::Tensor & out, const at::Tensor & self, ::std::optional<int64_t> dense_dim=::std::nullopt);
+TORCH_API at::Tensor & _to_sparse_csr_outf(const at::Tensor & self, ::std::optional<int64_t> dense_dim, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_semi_structured_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_semi_structured_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..969b51cd7ed42a95b07cfdb73b6e951323a29804
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_to_sparse_semi_structured_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> _to_sparse_semi_structured(const at::Tensor & dense);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_transformer_encoder_layer_fwd_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_transformer_encoder_layer_fwd_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d786066d695ce41a7bea9bce3c95b634c03cb6b1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_transformer_encoder_layer_fwd_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & _transformer_encoder_layer_fwd_out(at::Tensor & out, const at::Tensor & src, int64_t embed_dim, int64_t num_heads, const at::Tensor & qkv_weight, const at::Tensor & qkv_bias, const at::Tensor & proj_weight, const at::Tensor & proj_bias, bool use_gelu, bool norm_first, double eps, const at::Tensor & norm_weight_1, const at::Tensor & norm_bias_1, const at::Tensor & norm_weight_2, const at::Tensor & norm_bias_2, const at::Tensor & ffn_weight_1, const at::Tensor & ffn_bias_1, const at::Tensor & ffn_weight_2, const at::Tensor & ffn_bias_2, const ::std::optional<at::Tensor> & mask={}, ::std::optional<int64_t> mask_type=::std::nullopt);
+TORCH_API at::Tensor & _transformer_encoder_layer_fwd_outf(const at::Tensor & src, int64_t embed_dim, int64_t num_heads, const at::Tensor & qkv_weight, const at::Tensor & qkv_bias, const at::Tensor & proj_weight, const at::Tensor & proj_bias, bool use_gelu, bool norm_first, double eps, const at::Tensor & norm_weight_1, const at::Tensor & norm_bias_1, const at::Tensor & norm_weight_2, const at::Tensor & norm_bias_2, const at::Tensor & ffn_weight_1, const at::Tensor & ffn_bias_1, const at::Tensor & ffn_weight_2, const at::Tensor & ffn_bias_2, const ::std::optional<at::Tensor> & mask, ::std::optional<int64_t> mask_type, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_triton_multi_head_attention_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_triton_multi_head_attention_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ebe2db9719d00aa0b049ac5c65d262d111104468
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_triton_multi_head_attention_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _triton_multi_head_attention(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, int64_t embed_dim, int64_t num_head, const at::Tensor & qkv_weight, const at::Tensor & qkv_bias, const at::Tensor & proj_weight, const at::Tensor & proj_bias, const ::std::optional<at::Tensor> & mask={});
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_triton_scaled_dot_attention_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_triton_scaled_dot_attention_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f897d253f8fdf2e7e383ba507967caf78eb90062
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_triton_scaled_dot_attention_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _triton_scaled_dot_attention(const at::Tensor & q, const at::Tensor & k, const at::Tensor & v, double dropout_p=0.0);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_unique2_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_unique2_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..1cfbc002872cedde251b370f9b5706df0a795cea
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_unique2_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _unique2(const at::Tensor & self, bool sorted=true, bool return_inverse=false, bool return_counts=false);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_unsafe_view_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_unsafe_view_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..b857496ebb9b1d2745f916ae8d20284639fc5c21
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_unsafe_view_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _unsafe_view {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_unsafe_view";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_unsafe_view(Tensor self, SymInt[] size) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymIntArrayRef size);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef size);
+};
+
+struct TORCH_API _unsafe_view_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymIntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_unsafe_view";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_unsafe_view.out(Tensor self, SymInt[] size, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymIntArrayRef size, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef size, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_bicubic2d_aa_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_bicubic2d_aa_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..0300e1f673106d97d40900676b505487f5628920
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_bicubic2d_aa_backward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _upsample_bicubic2d_aa_backward_grad_input {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, ::std::optional<double>, ::std::optional<double>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_upsample_bicubic2d_aa_backward";
+  static constexpr const char* overload_name = "grad_input";
+  static constexpr const char* schema_str = "_upsample_bicubic2d_aa_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!)";
+  static at::Tensor & call(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);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, 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);
+};
+
+struct TORCH_API _upsample_bicubic2d_aa_backward {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, ::std::optional<double>, ::std::optional<double>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_upsample_bicubic2d_aa_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_upsample_bicubic2d_aa_backward(Tensor grad_output, SymInt[2] output_size, SymInt[4] input_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor";
+  static at::Tensor call(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);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, 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);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_bilinear2d_aa_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_bilinear2d_aa_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f030b570f6f41ce6163ef0fa1fe5ed99edd853e8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_bilinear2d_aa_cuda_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_bilinear2d_aa(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor _upsample_bilinear2d_aa_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & _upsample_bilinear2d_aa_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & _upsample_bilinear2d_aa_outf(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+TORCH_API at::Tensor & _upsample_bilinear2d_aa_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & _upsample_bilinear2d_aa_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e417f849c109aafcca7ea398f4159766c302ef23
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact2d_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact2d_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..7726d2920f4f7601c767c6daefdad3464ef2c980
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact2d_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _upsample_nearest_exact2d(const at::Tensor & input, at::OptionalIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors);
+TORCH_API at::Tensor _upsample_nearest_exact2d_symint(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact2d_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact2d_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..59e12db7b6b9a49fc86e8ee07d30898b312beff7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact2d_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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__upsample_nearest_exact2d : public at::impl::MetaBase {
+
+
+    void meta(const at::Tensor & self, at::ArrayRef<int64_t> output_size, ::std::optional<double> scales_h, ::std::optional<double> scales_w);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_backward_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_backward_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6dc7c439c67010d92ef471ad7a4c82f263b87f6b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_backward_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_exact3d_backward(const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_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_backward_symint(const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, ::std::optional<double> scales_d=::std::nullopt, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..1098ba4118de870d0b08a8859508fa9192dc4550
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _upsample_nearest_exact3d_vec {
+  using schema = at::Tensor (const at::Tensor &, at::OptionalSymIntArrayRef, ::std::optional<at::ArrayRef<double>>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_upsample_nearest_exact3d";
+  static constexpr const char* overload_name = "vec";
+  static constexpr const char* schema_str = "_upsample_nearest_exact3d.vec(Tensor input, SymInt[]? output_size, float[]? scale_factors) -> Tensor";
+  static at::Tensor call(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, at::OptionalSymIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors);
+};
+
+struct TORCH_API _upsample_nearest_exact3d_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymIntArrayRef, ::std::optional<double>, ::std::optional<double>, ::std::optional<double>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_upsample_nearest_exact3d";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "_upsample_nearest_exact3d.out(Tensor self, SymInt[3] output_size, float? scales_d=None, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(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);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, 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);
+};
+
+struct TORCH_API _upsample_nearest_exact3d {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef, ::std::optional<double>, ::std::optional<double>, ::std::optional<double>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_upsample_nearest_exact3d";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_upsample_nearest_exact3d(Tensor self, SymInt[3] output_size, float? scales_d=None, float? scales_h=None, float? scales_w=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales_d, ::std::optional<double> scales_h, ::std::optional<double> scales_w);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales_d, ::std::optional<double> scales_h, ::std::optional<double> scales_w);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_compressed_sparse_indices.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_compressed_sparse_indices.h
new file mode 100644
index 0000000000000000000000000000000000000000..9c3da14d5839e5868b096fed6f4d282c61d5885b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_compressed_sparse_indices.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_validate_compressed_sparse_indices_ops.h>
+
+namespace at {
+
+
+// aten::_validate_compressed_sparse_indices(bool is_crow, Tensor compressed_idx, Tensor plain_idx, int cdim, int dim, int nnz) -> ()
+inline void _validate_compressed_sparse_indices(bool is_crow, const at::Tensor & compressed_idx, const at::Tensor & plain_idx, int64_t cdim, int64_t dim, int64_t nnz) {
+    return at::_ops::_validate_compressed_sparse_indices::call(is_crow, compressed_idx, plain_idx, cdim, dim, nnz);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_compressed_sparse_indices_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_compressed_sparse_indices_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..8ec5be86aff469f10ceb0eb9c335f5f6513e3ca2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_compressed_sparse_indices_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void _validate_compressed_sparse_indices_cpu(bool is_crow, const at::Tensor & compressed_idx, const at::Tensor & plain_idx, int64_t cdim, int64_t dim, int64_t nnz);
+TORCH_API void _validate_compressed_sparse_indices_cuda(bool is_crow, const at::Tensor & compressed_idx, const at::Tensor & plain_idx, int64_t cdim, int64_t dim, int64_t nnz);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_sparse_compressed_tensor_args_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_sparse_compressed_tensor_args_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b217485dc0eadd1ab02487f3e292c57bb92788a9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_sparse_compressed_tensor_args_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void _validate_sparse_compressed_tensor_args(const at::Tensor & compressed_indices, const at::Tensor & plain_indices, const at::Tensor & values, at::IntArrayRef size, at::Layout layout, ::std::optional<bool> check_pinning=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_sparse_csc_tensor_args_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_sparse_csc_tensor_args_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..6460fb70684aec838e2e858201916df9c3842eb9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_sparse_csc_tensor_args_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 _validate_sparse_csc_tensor_args {
+  using schema = void (const at::Tensor &, const at::Tensor &, const at::Tensor &, at::IntArrayRef, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::_validate_sparse_csc_tensor_args";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "_validate_sparse_csc_tensor_args(Tensor ccol_indices, Tensor row_indices, Tensor values, int[] size, bool? check_pinning=None) -> ()";
+  static void call(const at::Tensor & ccol_indices, const at::Tensor & row_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<bool> check_pinning);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & ccol_indices, const at::Tensor & row_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<bool> check_pinning);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_sparse_csr_tensor_args_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_sparse_csr_tensor_args_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5a47fc3bc1a033ae469120313b19ca0b9525e548
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_validate_sparse_csr_tensor_args_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void _validate_sparse_csr_tensor_args(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<bool> check_pinning=::std::nullopt);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_version.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_version.h
new file mode 100644
index 0000000000000000000000000000000000000000..009c4295100cb293cc19fefc0edbec8926417e33
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_version.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/_version_ops.h>
+
+namespace at {
+
+
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_version_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_version_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ca668d18900e7fa6876e928f0f381d7be4151b52
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_version_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API int64_t _version(const at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_int4pack_mm_for_cpu_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_int4pack_mm_for_cpu_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..361f631f9496d76a322c2f5dd74585fbbdf29969
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_int4pack_mm_for_cpu_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 _weight_int4pack_mm_for_cpu(const at::Tensor & self, const at::Tensor & mat2, int64_t qGroupSize, const at::Tensor & qScaleAndZeros);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_norm_interface_backward_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_norm_interface_backward_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..0cec805bf9505020bb7cea7091a66af3b7ea667a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_norm_interface_backward_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 &> _weight_norm_interface_backward_out(at::Tensor & out0, at::Tensor & out1, const at::Tensor & grad_w, const at::Tensor & saved_v, const at::Tensor & saved_g, const at::Tensor & saved_norms, int64_t dim);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> _weight_norm_interface_backward_outf(const at::Tensor & grad_w, const at::Tensor & saved_v, const at::Tensor & saved_g, const at::Tensor & saved_norms, int64_t dim, at::Tensor & out0, at::Tensor & out1);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_norm_interface_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_norm_interface_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..9f2401ac35082d0798cc3ec6b34a9babdeecee25
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_norm_interface_backward_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> _weight_norm_interface_backward_out(const at::Tensor & grad_w, const at::Tensor & saved_v, const at::Tensor & saved_g, const at::Tensor & saved_norms, int64_t dim, at::Tensor & out0, at::Tensor & out1);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> weight_norm_backward_cpu(const at::Tensor & grad_w, const at::Tensor & saved_v, const at::Tensor & saved_g, const at::Tensor & saved_norms, int64_t dim);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> weight_norm_backward_cuda(const at::Tensor & grad_w, const at::Tensor & saved_v, const at::Tensor & saved_g, const at::Tensor & saved_norms, int64_t dim);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_norm_interface_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_norm_interface_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b7ab1d2479f34e8e2833fe335995fef4925aa904
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_weight_norm_interface_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> _weight_norm_interface(const at::Tensor & v, const at::Tensor & g, int64_t dim=0);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_wrapped_quantized_linear_prepacked_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_wrapped_quantized_linear_prepacked_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a81eb440a7c68263c87ddfe2aa330d98a967971c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/_wrapped_quantized_linear_prepacked_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor _wrapped_quantized_linear_prepacked(const at::Tensor & input, const at::Tensor & input_scale, const at::Tensor & input_zero_point, const at::Tensor & packed_weight, const at::Tensor & output_scale, const at::Tensor & output_zero_point, int64_t out_channel);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/acos_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/acos_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..d99e88b7677af1fa253c6e76eb63745d7334904f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/acos_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/acos_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_acos_out : public at::meta::structured_acos {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_avg_pool1d_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_avg_pool1d_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..1deca48673ee73e2e578389928d48cc2866aa6d3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_avg_pool1d_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 adaptive_avg_pool1d(const at::Tensor & self, at::IntArrayRef output_size);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_avg_pool1d_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_avg_pool1d_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..a04d9c7ef945d841f22d96e901d9321d26900cc5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_avg_pool1d_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 adaptive_avg_pool1d(const at::Tensor & self, at::IntArrayRef output_size);
+TORCH_API at::Tensor & adaptive_avg_pool1d_out(const at::Tensor & self, at::IntArrayRef output_size, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_avg_pool1d_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_avg_pool1d_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..962c6389d9713d4a06b5ab245dc339848a2ff4e1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_avg_pool1d_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 adaptive_avg_pool1d {
+  using schema = at::Tensor (const at::Tensor &, at::IntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::adaptive_avg_pool1d";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "adaptive_avg_pool1d(Tensor self, int[1] output_size) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::IntArrayRef output_size);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef output_size);
+};
+
+struct TORCH_API adaptive_avg_pool1d_out {
+  using schema = at::Tensor & (const at::Tensor &, at::IntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::adaptive_avg_pool1d";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "adaptive_avg_pool1d.out(Tensor self, int[1] output_size, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, at::IntArrayRef output_size, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef output_size, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool2d_backward_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool2d_backward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..11e1ae5cd8f575ac2d6cbdcf0a52b75702c9ff84
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool2d_backward_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor adaptive_max_pool2d_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices);
+TORCH_API at::Tensor & adaptive_max_pool2d_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices);
+TORCH_API at::Tensor & adaptive_max_pool2d_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices, at::Tensor & grad_input);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_backward_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_backward_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5262359de7788dbca32a7d8c6f1141e017c9c490
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_backward_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 adaptive_max_pool3d_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_backward_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_backward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a45cff56658a4e61357fd0b9930d3fbb86a39cff
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_backward_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor adaptive_max_pool3d_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices);
+TORCH_API at::Tensor & adaptive_max_pool3d_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices);
+TORCH_API at::Tensor & adaptive_max_pool3d_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices, at::Tensor & grad_input);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..b02249f6af1e17cf537c0918367e952f73367e4a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/adaptive_max_pool3d_backward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 adaptive_max_pool3d_backward_grad_input {
+  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 const char* name = "aten::adaptive_max_pool3d_backward";
+  static constexpr const char* overload_name = "grad_input";
+  static constexpr const char* schema_str = "adaptive_max_pool3d_backward.grad_input(Tensor grad_output, Tensor self, Tensor indices, *, Tensor(a!) grad_input) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices, at::Tensor & grad_input);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices, at::Tensor & grad_input);
+};
+
+struct TORCH_API adaptive_max_pool3d_backward {
+  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 const char* name = "aten::adaptive_max_pool3d_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "adaptive_max_pool3d_backward(Tensor grad_output, Tensor self, Tensor indices) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/add_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/add_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..560cab5df9227e5b68ce0b12feec4ebe4f6604fa
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/add_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 add(const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & add_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & add_outf(const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & add_(at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha=1);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/add_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/add_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8a244bf1c016bb3a4d069e182236359ae766a2a4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/add_ops.h
@@ -0,0 +1,89 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 add_Tensor {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::add";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "add.Tensor(Tensor self, Tensor other, *, Scalar alpha=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha);
+};
+
+struct TORCH_API add__Tensor {
+  using schema = at::Tensor & (at::Tensor &, const at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::add_";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "add_.Tensor(Tensor(a!) self, Tensor other, *, Scalar alpha=1) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha);
+};
+
+struct TORCH_API add_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::add";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "add.out(Tensor self, Tensor other, *, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out);
+};
+
+struct TORCH_API add_Scalar {
+  using schema = at::Tensor (const at::Tensor &, const at::Scalar &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::add";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "add.Scalar(Tensor self, Scalar other, Scalar alpha=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha);
+};
+
+struct TORCH_API add__Scalar {
+  using schema = at::Tensor & (at::Tensor &, const at::Scalar &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::add_";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "add_.Scalar(Tensor(a!) self, Scalar other, Scalar alpha=1) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha);
+};
+
+struct TORCH_API add_Scalar_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Scalar &, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::add";
+  static constexpr const char* overload_name = "Scalar_out";
+  static constexpr const char* schema_str = "add.Scalar_out(Tensor self, Scalar other, Scalar alpha=1, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmm_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmm_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4a66257846f75aeb0244fffda26779903ec99694
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmm_cuda_dispatch.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 addmm(const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addmm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addmm_outf(const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & addmm_(at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor addmm(const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, at::ScalarType out_dtype, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addmm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, at::ScalarType out_dtype, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addmm_outf(const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, at::ScalarType out_dtype, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv.h
new file mode 100644
index 0000000000000000000000000000000000000000..5cbaf9bcae1c4d5f48da1eb976dc4bbc9e377a1c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/addmv_ops.h>
+
+namespace at {
+
+
+// aten::addmv(Tensor self, Tensor mat, Tensor vec, *, Scalar beta=1, Scalar alpha=1) -> Tensor
+inline at::Tensor addmv(const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta=1, const at::Scalar & alpha=1) {
+    return at::_ops::addmv::call(self, mat, vec, beta, alpha);
+}
+
+// aten::addmv_(Tensor(a!) self, Tensor mat, Tensor vec, *, Scalar beta=1, Scalar alpha=1) -> Tensor(a!)
+inline at::Tensor & addmv_(at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta=1, const at::Scalar & alpha=1) {
+    return at::_ops::addmv_::call(self, mat, vec, beta, alpha);
+}
+
+// aten::addmv.out(Tensor self, Tensor mat, Tensor vec, *, Scalar beta=1, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & addmv_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta=1, const at::Scalar & alpha=1) {
+    return at::_ops::addmv_out::call(self, mat, vec, beta, alpha, out);
+}
+// aten::addmv.out(Tensor self, Tensor mat, Tensor vec, *, Scalar beta=1, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & addmv_outf(const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out) {
+    return at::_ops::addmv_out::call(self, mat, vec, beta, alpha, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b7918d99300b591eb3707a562134751135364696
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 addmv(const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addmv_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addmv_outf(const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & addmv_(at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a5a7de8e90bc1014950dd63921226a4b7d4c07f7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 addmv(const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addmv_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addmv_outf(const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & addmv_(at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..c56ab77c8c1cf46119d754864b23007477a99fec
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addmv_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 addmv {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Scalar &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::addmv";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "addmv(Tensor self, Tensor mat, Tensor vec, *, Scalar beta=1, Scalar alpha=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta, const at::Scalar & alpha);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta, const at::Scalar & alpha);
+};
+
+struct TORCH_API addmv_ {
+  using schema = at::Tensor & (at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Scalar &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::addmv_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "addmv_(Tensor(a!) self, Tensor mat, Tensor vec, *, Scalar beta=1, Scalar alpha=1) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta, const at::Scalar & alpha);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta, const at::Scalar & alpha);
+};
+
+struct TORCH_API addmv_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Scalar &, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::addmv";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "addmv.out(Tensor self, Tensor mat, Tensor vec, *, Scalar beta=1, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat, const at::Tensor & vec, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addr_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addr_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3cfce64203ce55e10145891889ac6c6ac5f89fd2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addr_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 addr(const at::Tensor & self, const at::Tensor & vec1, const at::Tensor & vec2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addr_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & vec1, const at::Tensor & vec2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addr_outf(const at::Tensor & self, const at::Tensor & vec1, const at::Tensor & vec2, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addr_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addr_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b5f25dcf6fe329aa139be9634809d3084a5ca773
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addr_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 addr(const at::Tensor & self, const at::Tensor & vec1, const at::Tensor & vec2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addr_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & vec1, const at::Tensor & vec2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addr_outf(const at::Tensor & self, const at::Tensor & vec1, const at::Tensor & vec2, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addr_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addr_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..08fe22421af7750a01c9e17cdf97f5aeeff931bb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/addr_native.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 math_addr(const at::Tensor & self, const at::Tensor & vec1, const at::Tensor & vec2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & math_addr_out(const at::Tensor & self, const at::Tensor & vec1, const at::Tensor & vec2, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & addr_(at::Tensor & self, const at::Tensor & vec1, const at::Tensor & vec2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor addr(const at::Tensor & self, const at::Tensor & vec1, const at::Tensor & vec2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addr_out(const at::Tensor & self, const at::Tensor & vec1, const at::Tensor & vec2, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/affine_grid_generator_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/affine_grid_generator_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..ffa01e4b7b61c1d5c5daa7452bfc042fe3943d0e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/affine_grid_generator_backward_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 affine_grid_generator_backward(const at::Tensor & grad, at::IntArrayRef size, bool align_corners);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/alias_copy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/alias_copy.h
new file mode 100644
index 0000000000000000000000000000000000000000..e1a60453a4478b00b70f42ac6154d8aa4c0930cd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/alias_copy.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/alias_copy_ops.h>
+
+namespace at {
+
+
+// aten::alias_copy(Tensor self) -> Tensor
+inline at::Tensor alias_copy(const at::Tensor & self) {
+    return at::_ops::alias_copy::call(self);
+}
+
+// aten::alias_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & alias_copy_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::alias_copy_out::call(self, out);
+}
+// aten::alias_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & alias_copy_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::alias_copy_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/alias_copy_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/alias_copy_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..27f6d37cb2608c909137e92cdbb1c2d3e04c3ece
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/alias_copy_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 alias_copy {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::alias_copy";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "alias_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 alias_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 const char* name = "aten::alias_copy";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "alias_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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/align_as_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/align_as_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..ae3d4f7c2d81b7f58dad4b09281d86e27ff655a8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/align_as_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 align_as(const at::Tensor & self, const at::Tensor & other);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/align_to_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/align_to_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..8328192c20150576121117e7b441da3fc2299101
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/align_to_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 align_to(const at::Tensor & self, at::DimnameList names);
+TORCH_API at::Tensor align_to(const at::Tensor & self, at::DimnameList order, int64_t ellipsis_idx);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/all_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/all_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c0342b10727e5db2bc6dbd2c46108f7dc6dc8371
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/all_meta_dispatch.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 all(const at::Tensor & self, int64_t dim, bool keepdim=false);
+TORCH_API at::Tensor & all_out(at::Tensor & out, const at::Tensor & self, int64_t dim, bool keepdim=false);
+TORCH_API at::Tensor & all_outf(const at::Tensor & self, int64_t dim, bool keepdim, at::Tensor & out);
+TORCH_API at::Tensor all(const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim=false);
+TORCH_API at::Tensor & all_out(at::Tensor & out, const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim=false);
+TORCH_API at::Tensor & all_outf(const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim, at::Tensor & out);
+TORCH_API at::Tensor all(const at::Tensor & self);
+TORCH_API at::Tensor & all_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & all_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/allclose_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/allclose_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..5ba5108bcb96f86e71e67b96b3641e38d921e713
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/allclose_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 allclose {
+  using schema = bool (const at::Tensor &, const at::Tensor &, double, double, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::allclose";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "allclose(Tensor self, Tensor other, float rtol=1e-05, float atol=1e-08, bool equal_nan=False) -> bool";
+  static bool call(const at::Tensor & self, const at::Tensor & other, double rtol, double atol, bool equal_nan);
+  static bool redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, double rtol, double atol, bool equal_nan);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/alpha_dropout_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/alpha_dropout_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..2839444c3ed0e8262b07d79a988e5533bc0d9141
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/alpha_dropout_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 alpha_dropout(const at::Tensor & input, double p, bool train);
+TORCH_API at::Tensor & alpha_dropout_(at::Tensor & self, double p, bool train);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/amax.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/amax.h
new file mode 100644
index 0000000000000000000000000000000000000000..807dbc68ba127ca5ad080f78cdfde84c19fde3a4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/amax.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/amax_ops.h>
+
+namespace at {
+
+
+// aten::amax(Tensor self, int[1] dim=[], bool keepdim=False) -> Tensor
+inline at::Tensor amax(const at::Tensor & self, at::IntArrayRef dim={}, bool keepdim=false) {
+    return at::_ops::amax::call(self, dim, keepdim);
+}
+
+// aten::amax.out(Tensor self, int[1] dim=[], bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & amax_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef dim={}, bool keepdim=false) {
+    return at::_ops::amax_out::call(self, dim, keepdim, out);
+}
+// aten::amax.out(Tensor self, int[1] dim=[], bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & amax_outf(const at::Tensor & self, at::IntArrayRef dim, bool keepdim, at::Tensor & out) {
+    return at::_ops::amax_out::call(self, dim, keepdim, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/amin_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/amin_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3227ac36f0a7a20044250b40fd10ebe9f525da2c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/amin_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 amin(const at::Tensor & self, at::IntArrayRef dim={}, bool keepdim=false);
+TORCH_API at::Tensor & amin_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef dim={}, bool keepdim=false);
+TORCH_API at::Tensor & amin_outf(const at::Tensor & self, at::IntArrayRef dim, bool keepdim, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/amin_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/amin_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..3666356a38a3756c2d418f92d54965f7320ab723
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/amin_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/amin_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_amin_out : public at::meta::structured_amin {
+void impl(const at::Tensor & self, at::IntArrayRef dim, bool keepdim, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/aminmax.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/aminmax.h
new file mode 100644
index 0000000000000000000000000000000000000000..09ca85ffe2803331c2c6bf3f3aa0627223ede533
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/aminmax.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/aminmax_ops.h>
+
+namespace at {
+
+
+// aten::aminmax(Tensor self, *, int? dim=None, bool keepdim=False) -> (Tensor min, Tensor max)
+inline ::std::tuple<at::Tensor,at::Tensor> aminmax(const at::Tensor & self, ::std::optional<int64_t> dim=::std::nullopt, bool keepdim=false) {
+    return at::_ops::aminmax::call(self, dim, keepdim);
+}
+
+// aten::aminmax.out(Tensor self, *, int? dim=None, bool keepdim=False, Tensor(a!) min, Tensor(b!) max) -> (Tensor(a!) min, Tensor(b!) max)
+inline ::std::tuple<at::Tensor &,at::Tensor &> aminmax_out(at::Tensor & min, at::Tensor & max, const at::Tensor & self, ::std::optional<int64_t> dim=::std::nullopt, bool keepdim=false) {
+    return at::_ops::aminmax_out::call(self, dim, keepdim, min, max);
+}
+// aten::aminmax.out(Tensor self, *, int? dim=None, bool keepdim=False, Tensor(a!) min, Tensor(b!) max) -> (Tensor(a!) min, Tensor(b!) max)
+inline ::std::tuple<at::Tensor &,at::Tensor &> aminmax_outf(const at::Tensor & self, ::std::optional<int64_t> dim, bool keepdim, at::Tensor & min, at::Tensor & max) {
+    return at::_ops::aminmax_out::call(self, dim, keepdim, min, max);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/aminmax_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/aminmax_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..b2f0f03aeafe38473ef1f98c754ba1d857c63778
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/aminmax_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_aminmax : public at::impl::MetaBase {
+
+
+    void meta(const at::Tensor & self, ::std::optional<int64_t> dim, bool keepdim);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/and_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/and_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9b45f48b98915a40c454268dd54b397f72bcc864
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/and_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 __and__(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & __iand__(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor __and__(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & __iand__(at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/angle_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/angle_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..4c96c857dd55c790e42882f0a76d9a7543872311
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/angle_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 angle(const at::Tensor & self);
+TORCH_API at::Tensor & angle_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor angle_sparse_csr(const at::Tensor & self);
+TORCH_API at::Tensor & angle_sparse_csr_out(const at::Tensor & self, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arange_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arange_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..d29004da740199ba87ee2593bcb8f9b07fd4e12e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arange_native.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 arange(const at::Scalar & end, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor & arange_out(const at::Scalar & end, at::Tensor & out);
+TORCH_API at::Tensor arange(const at::Scalar & start, const at::Scalar & end, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor arange(const at::Scalar & start, const at::Scalar & end, const at::Scalar & step=1, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor & arange_out(const at::Scalar & start, const at::Scalar & end, const at::Scalar & step, at::Tensor & out);
+TORCH_API at::Tensor & arange_cuda_out(const at::Scalar & start, const at::Scalar & end, const at::Scalar & step, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arccos_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arccos_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..9f35ded4a53270e5613496b9856f4ebd33224c05
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arccos_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 arccos(const at::Tensor & self);
+TORCH_API at::Tensor & arccos_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & arccos_(at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arccosh.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arccosh.h
new file mode 100644
index 0000000000000000000000000000000000000000..4c303ef5e1909b212d5ec4e0a94046a4cf4dabfa
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arccosh.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/arccosh_ops.h>
+
+namespace at {
+
+
+// aten::arccosh(Tensor self) -> Tensor
+inline at::Tensor arccosh(const at::Tensor & self) {
+    return at::_ops::arccosh::call(self);
+}
+
+// aten::arccosh_(Tensor(a!) self) -> Tensor(a!)
+inline at::Tensor & arccosh_(at::Tensor & self) {
+    return at::_ops::arccosh_::call(self);
+}
+
+// aten::arccosh.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & arccosh_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::arccosh_out::call(self, out);
+}
+// aten::arccosh.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & arccosh_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::arccosh_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arccosh_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arccosh_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b269b8cbfcbbae75dbc1e3d2c498837b113f0307
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arccosh_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 arccosh(const at::Tensor & self);
+TORCH_API at::Tensor & arccosh_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & arccosh_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & arccosh_(at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsin_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsin_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4fcba54cdfc93703a326cf9c4cba3141bce37294
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsin_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 arcsin(const at::Tensor & self);
+TORCH_API at::Tensor & arcsin_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & arcsin_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & arcsin_(at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsin_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsin_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..9c59ea88c30befea08aaa4922182a78a1c609c9e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsin_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 arcsin {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::arcsin";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "arcsin(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 arcsin_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::arcsin_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "arcsin_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API arcsin_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 const char* name = "aten::arcsin";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "arcsin.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsinh_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsinh_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..137da143773699330c35ef818857487ed9f078bf
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsinh_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 arcsinh(const at::Tensor & self);
+TORCH_API at::Tensor & arcsinh_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & arcsinh_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & arcsinh_(at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsinh_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsinh_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..2f5fdefd337847ab7494edd773195b18c8ac218f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arcsinh_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 arcsinh {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::arcsinh";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "arcsinh(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 arcsinh_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::arcsinh_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "arcsinh_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API arcsinh_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 const char* name = "aten::arcsinh";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "arcsinh.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arctan_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arctan_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..496cd80495cc3c91f3d2afa2c0ac2e0616660205
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arctan_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 arctan(const at::Tensor & self);
+TORCH_API at::Tensor & arctan_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & arctan_(at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arctan_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arctan_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..c5b92b61ac3e8cff58ba115a066887f2ffa8f951
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/arctan_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 arctan {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::arctan";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "arctan(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 arctan_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::arctan_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "arctan_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API arctan_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 const char* name = "aten::arctan";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "arctan.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/argmin_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/argmin_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..deabe1cdf272838f4f3f1bf45c1d289964bd16f7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/argmin_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/argmin_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_argmin_out : public at::meta::structured_argmin {
+void impl(const at::Tensor & self, ::std::optional<int64_t> dim, bool keepdim, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/argwhere_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/argwhere_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..2b5c0b3f84e5726befdbd7499508e94384a0c328
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/argwhere_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 argwhere {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::argwhere";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "argwhere(Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided.h
new file mode 100644
index 0000000000000000000000000000000000000000..43f9048d6a7653e6be53f91ea6ca18014295d8e4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided.h
@@ -0,0 +1,75 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/as_strided_ops.h>
+
+namespace at {
+
+
+// aten::as_strided(Tensor(a) self, SymInt[] size, SymInt[] stride, SymInt? storage_offset=None) -> Tensor(a)
+inline at::Tensor as_strided(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<int64_t> storage_offset=::std::nullopt) {
+    return at::_ops::as_strided::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride), storage_offset.has_value() ? ::std::make_optional(c10::SymInt(*storage_offset)) : ::std::nullopt);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor as_strided(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<int64_t> storage_offset=::std::nullopt) {
+    return at::_ops::as_strided::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride), storage_offset.has_value() ? ::std::make_optional(c10::SymInt(*storage_offset)) : ::std::nullopt);
+  }
+}
+
+// aten::as_strided(Tensor(a) self, SymInt[] size, SymInt[] stride, SymInt? storage_offset=None) -> Tensor(a)
+inline at::Tensor as_strided_symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset=::std::nullopt) {
+    return at::_ops::as_strided::call(self, size, stride, storage_offset);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor as_strided(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset=::std::nullopt) {
+    return at::_ops::as_strided::call(self, size, stride, storage_offset);
+  }
+}
+
+// aten::as_strided_(Tensor(a!) self, SymInt[] size, SymInt[] stride, SymInt? storage_offset=None) -> Tensor(a!)
+inline const at::Tensor & as_strided_(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<int64_t> storage_offset=::std::nullopt) {
+    return at::_ops::as_strided_::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride), storage_offset.has_value() ? ::std::make_optional(c10::SymInt(*storage_offset)) : ::std::nullopt);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  const at::Tensor & as_strided_(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<int64_t> storage_offset=::std::nullopt) {
+    return at::_ops::as_strided_::call(self, c10::fromIntArrayRefSlow(size), c10::fromIntArrayRefSlow(stride), storage_offset.has_value() ? ::std::make_optional(c10::SymInt(*storage_offset)) : ::std::nullopt);
+  }
+}
+
+// aten::as_strided_(Tensor(a!) self, SymInt[] size, SymInt[] stride, SymInt? storage_offset=None) -> Tensor(a!)
+inline const at::Tensor & as_strided__symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset=::std::nullopt) {
+    return at::_ops::as_strided_::call(self, size, stride, storage_offset);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  const at::Tensor & as_strided_(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset=::std::nullopt) {
+    return at::_ops::as_strided_::call(self, size, stride, storage_offset);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_copy_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_copy_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f66c725027c96e0ccf657256d3ef171ea5162b18
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_copy_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & as_strided_copy_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<int64_t> storage_offset=::std::nullopt);
+TORCH_API at::Tensor & as_strided_copy_outf(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<int64_t> storage_offset, at::Tensor & out);
+TORCH_API at::Tensor & as_strided_copy_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset=::std::nullopt);
+TORCH_API at::Tensor & as_strided_copy_symint_outf(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_copy_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_copy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..69e68b7514baa3ca3748e08fd10cd94022b4a043
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_copy_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & as_strided_copy_out_symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset, at::Tensor & out);
+TORCH_API at::Tensor as_strided_copy_symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_copy_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_copy_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..c877e57aa06a225dc73910db8d2ac3c420bee8e3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_copy_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 as_strided_copy {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, ::std::optional<c10::SymInt>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::as_strided_copy";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "as_strided_copy(Tensor self, SymInt[] size, SymInt[] stride, SymInt? storage_offset=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset);
+};
+
+struct TORCH_API as_strided_copy_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, ::std::optional<c10::SymInt>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::as_strided_copy";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "as_strided_copy.out(Tensor self, SymInt[] size, SymInt[] stride, SymInt? storage_offset=None, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..da44a5c83649acf488b5b9f5b9b38430f50d7639
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 as_strided_tensorimpl(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<int64_t> storage_offset=::std::nullopt);
+TORCH_API at::Tensor as_strided_tensorimpl_meta_symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset=::std::nullopt);
+TORCH_API at::Tensor as_strided_qtensorimpl(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<int64_t> storage_offset=::std::nullopt);
+TORCH_API const at::Tensor & as_strided__symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_scatter_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_scatter_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6cb5a721773e522b779c423db8b803e16c954472
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_scatter_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 as_strided_scatter(const at::Tensor & self, const at::Tensor & src, at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<int64_t> storage_offset=::std::nullopt);
+TORCH_API at::Tensor as_strided_scatter_symint(const at::Tensor & self, const at::Tensor & src, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset=::std::nullopt);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_scatter_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_scatter_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..9642f80b4b838d462473581a6b7b6b14d582d95f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/as_strided_scatter_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 as_strided_scatter {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, ::std::optional<c10::SymInt>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::as_strided_scatter";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "as_strided_scatter(Tensor self, Tensor src, SymInt[] size, SymInt[] stride, SymInt? storage_offset=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & src, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & src, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset);
+};
+
+struct TORCH_API as_strided_scatter_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, ::std::optional<c10::SymInt>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::as_strided_scatter";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "as_strided_scatter.out(Tensor self, Tensor src, SymInt[] size, SymInt[] stride, SymInt? storage_offset=None, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & src, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & src, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asin.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asin.h
new file mode 100644
index 0000000000000000000000000000000000000000..9f03ee48130404337099e64f837008a898b9e9d3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asin.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/asin_ops.h>
+
+namespace at {
+
+
+// aten::asin(Tensor self) -> Tensor
+inline at::Tensor asin(const at::Tensor & self) {
+    return at::_ops::asin::call(self);
+}
+
+// aten::asin_(Tensor(a!) self) -> Tensor(a!)
+inline at::Tensor & asin_(at::Tensor & self) {
+    return at::_ops::asin_::call(self);
+}
+
+// aten::asin.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & asin_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::asin_out::call(self, out);
+}
+// aten::asin.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & asin_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::asin_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asin_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asin_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9afcaa70ffbd8a5da9a2ed3f3aee15c08a07dc68
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asin_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 asin(const at::Tensor & self);
+TORCH_API at::Tensor & asin_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4e435b6809166ba60a54cf2c8635c06e069c8f52
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 asinh(const at::Tensor & self);
+TORCH_API at::Tensor & asinh_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & asinh_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & asinh_(at::Tensor & self);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..da7b37bddcff9b08de44343b1eaacd0bddccd6f1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 asinh(const at::Tensor & self);
+TORCH_API at::Tensor & asinh_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & asinh_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & asinh_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..89d10c4042914f63caa58725feb42d7888fa5c44
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_native.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/asinh_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_asinh_out : public at::meta::structured_asinh {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+TORCH_API at::Tensor asinh_sparse(const at::Tensor & self);
+TORCH_API at::Tensor & asinh_sparse_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & asinh_sparse_(at::Tensor & self);
+TORCH_API at::Tensor asinh_sparse_csr(const at::Tensor & self);
+TORCH_API at::Tensor & asinh_sparse_csr_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & asinh_sparse_csr_(at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..90ee4de61c6f06f95927ef9ab49343a3dc124a18
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/asinh_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 asinh {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::asinh";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "asinh(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 asinh_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::asinh_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "asinh_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API asinh_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 const char* name = "aten::asinh";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "asinh.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan.h
new file mode 100644
index 0000000000000000000000000000000000000000..bfae5e8938a1835ef501168a380862b905f47eb7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/atan_ops.h>
+
+namespace at {
+
+
+// aten::atan(Tensor self) -> Tensor
+inline at::Tensor atan(const at::Tensor & self) {
+    return at::_ops::atan::call(self);
+}
+
+// aten::atan_(Tensor(a!) self) -> Tensor(a!)
+inline at::Tensor & atan_(at::Tensor & self) {
+    return at::_ops::atan_::call(self);
+}
+
+// aten::atan.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & atan_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::atan_out::call(self, out);
+}
+// aten::atan.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & atan_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::atan_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan2_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan2_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..7551e659b8b9277bdbbcfe464ff03986629b8408
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan2_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 atan2(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & atan2_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & atan2_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & atan2_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan2_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan2_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..badaf57df00a4c1086e86fd4186d5bb4b6af0e92
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan2_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 atan2(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & atan2_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & atan2_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & atan2_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d4f65971ed4ca002edaf6c4fc0ae13e650532b55
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atan_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 atan(const at::Tensor & self);
+TORCH_API at::Tensor & atan_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & atan_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & atan_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atanh_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atanh_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c935f41d82f6c5d5e27e6bd57dc1efc81fdee69d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atanh_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 atanh(const at::Tensor & self);
+TORCH_API at::Tensor & atanh_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & atanh_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & atanh_(at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_1d.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_1d.h
new file mode 100644
index 0000000000000000000000000000000000000000..53e21c3b7b50ecba0bd07526d87c848e0bf9d14d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_1d.h
@@ -0,0 +1,41 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/atleast_1d_ops.h>
+
+namespace at {
+
+
+// aten::atleast_1d(Tensor self) -> Tensor
+inline at::Tensor atleast_1d(const at::Tensor & self) {
+    return at::_ops::atleast_1d::call(self);
+}
+
+// aten::atleast_1d.Sequence(Tensor[] tensors) -> Tensor[]
+inline ::std::vector<at::Tensor> atleast_1d(at::TensorList tensors) {
+    return at::_ops::atleast_1d_Sequence::call(tensors);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_1d_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_1d_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..0057ff3f832a4da2ab6f7084bbac97f856956f2a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_1d_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 atleast_1d(const at::Tensor & self);
+TORCH_API ::std::vector<at::Tensor> atleast_1d(at::TensorList tensors);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_2d_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_2d_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..deaf7b671db6d033a68111ea87f0daa367d468c6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_2d_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 atleast_2d(const at::Tensor & self);
+TORCH_API ::std::vector<at::Tensor> atleast_2d(at::TensorList tensors);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_2d_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_2d_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..0d9fe4fc860d70ca87366c3e11c109656d7a0b08
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/atleast_2d_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 atleast_2d {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::atleast_2d";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "atleast_2d(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 atleast_2d_Sequence {
+  using schema = ::std::vector<at::Tensor> (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::atleast_2d";
+  static constexpr const char* overload_name = "Sequence";
+  static constexpr const char* schema_str = "atleast_2d.Sequence(Tensor[] tensors) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList tensors);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList tensors);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool1d.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool1d.h
new file mode 100644
index 0000000000000000000000000000000000000000..ec0b2d448a33e0ce31ab56ab599e46280cf9a317
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool1d.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/avg_pool1d_ops.h>
+
+namespace at {
+
+
+// aten::avg_pool1d(Tensor self, int[1] kernel_size, int[1] stride=[], int[1] padding=0, bool ceil_mode=False, bool count_include_pad=True) -> Tensor
+inline at::Tensor avg_pool1d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, bool ceil_mode=false, bool count_include_pad=true) {
+    return at::_ops::avg_pool1d::call(self, kernel_size, stride, padding, ceil_mode, count_include_pad);
+}
+
+// aten::avg_pool1d.out(Tensor self, int[1] kernel_size, int[1] stride=[], int[1] padding=0, bool ceil_mode=False, bool count_include_pad=True, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & avg_pool1d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, bool ceil_mode=false, bool count_include_pad=true) {
+    return at::_ops::avg_pool1d_out::call(self, kernel_size, stride, padding, ceil_mode, count_include_pad, out);
+}
+// aten::avg_pool1d.out(Tensor self, int[1] kernel_size, int[1] stride=[], int[1] padding=0, bool ceil_mode=False, bool count_include_pad=True, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & avg_pool1d_outf(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, bool ceil_mode, bool count_include_pad, at::Tensor & out) {
+    return at::_ops::avg_pool1d_out::call(self, kernel_size, stride, padding, ceil_mode, count_include_pad, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool2d_backward_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool2d_backward_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f7006692c81da7f3124951ba3ad1526ac0be59ae
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool2d_backward_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 avg_pool2d_backward(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, bool ceil_mode, bool count_include_pad, ::std::optional<int64_t> divisor_override);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool2d_backward_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool2d_backward_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..e4e67ea2d47e0e55fc268455465298d5f3938b45
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool2d_backward_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_avg_pool2d_backward : public at::impl::MetaBase {
+
+
+    void meta(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, bool ceil_mode, bool count_include_pad, ::std::optional<int64_t> divisor_override);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d.h
new file mode 100644
index 0000000000000000000000000000000000000000..59ac86eacefe258c0ca8c6a9047e7286360b2cb4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/avg_pool3d_ops.h>
+
+namespace at {
+
+
+// aten::avg_pool3d.out(Tensor self, int[3] kernel_size, int[3] stride=[], int[3] padding=0, bool ceil_mode=False, bool count_include_pad=True, int? divisor_override=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & avg_pool3d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, bool ceil_mode=false, bool count_include_pad=true, ::std::optional<int64_t> divisor_override=::std::nullopt) {
+    return at::_ops::avg_pool3d_out::call(self, kernel_size, stride, padding, ceil_mode, count_include_pad, divisor_override, out);
+}
+// aten::avg_pool3d.out(Tensor self, int[3] kernel_size, int[3] stride=[], int[3] padding=0, bool ceil_mode=False, bool count_include_pad=True, int? divisor_override=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & avg_pool3d_outf(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, bool ceil_mode, bool count_include_pad, ::std::optional<int64_t> divisor_override, at::Tensor & out) {
+    return at::_ops::avg_pool3d_out::call(self, kernel_size, stride, padding, ceil_mode, count_include_pad, divisor_override, out);
+}
+
+// aten::avg_pool3d(Tensor self, int[3] kernel_size, int[3] stride=[], int[3] padding=0, bool ceil_mode=False, bool count_include_pad=True, int? divisor_override=None) -> Tensor
+inline at::Tensor avg_pool3d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, bool ceil_mode=false, bool count_include_pad=true, ::std::optional<int64_t> divisor_override=::std::nullopt) {
+    return at::_ops::avg_pool3d::call(self, kernel_size, stride, padding, ceil_mode, count_include_pad, divisor_override);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..55ec952496d49c1206d04ada0916f93eb17b019a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 avg_pool3d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, bool ceil_mode=false, bool count_include_pad=true, ::std::optional<int64_t> divisor_override=::std::nullopt);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..335df8354423966ceea626bfdfdadc7c786ad641
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 avg_pool3d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, bool ceil_mode=false, bool count_include_pad=true, ::std::optional<int64_t> divisor_override=::std::nullopt);
+TORCH_API at::Tensor & avg_pool3d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, bool ceil_mode=false, bool count_include_pad=true, ::std::optional<int64_t> divisor_override=::std::nullopt);
+TORCH_API at::Tensor & avg_pool3d_outf(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, bool ceil_mode, bool count_include_pad, ::std::optional<int64_t> divisor_override, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8571653bd24de62969e9cb05c9a5ba61cc578a9f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/avg_pool3d_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 avg_pool3d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, bool ceil_mode=false, bool count_include_pad=true, ::std::optional<int64_t> divisor_override=::std::nullopt);
+TORCH_API at::Tensor & avg_pool3d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, bool ceil_mode=false, bool count_include_pad=true, ::std::optional<int64_t> divisor_override=::std::nullopt);
+TORCH_API at::Tensor & avg_pool3d_outf(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, bool ceil_mode, bool count_include_pad, ::std::optional<int64_t> divisor_override, at::Tensor & out);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..405093a5bd51a2ebd59a233d3b27766650af2cb2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> batch_norm_backward(const at::Tensor & grad_out, const at::Tensor & input, const 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_var, bool update, double eps, ::std::array<bool,3> output_mask, const at::Tensor & reserve);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_elemt_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_elemt_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b72ec59b28c2d40918bf4b76fa5d0bef4bff59f6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_elemt_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & batch_norm_backward_elemt_out(at::Tensor & out, const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & invstd, const ::std::optional<at::Tensor> & weight, const at::Tensor & sum_dy, const at::Tensor & sum_dy_xmu, const at::Tensor & count);
+TORCH_API at::Tensor & batch_norm_backward_elemt_outf(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & invstd, const ::std::optional<at::Tensor> & weight, const at::Tensor & sum_dy, const at::Tensor & sum_dy_xmu, const at::Tensor & count, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_elemt_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_elemt_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..7aaad8aac0dfd034b1e4b0450c9906ad96ccb911
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_elemt_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & batch_norm_backward_elemt_out(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & invstd, const ::std::optional<at::Tensor> & weight, const at::Tensor & sum_dy, const at::Tensor & sum_dy_xmu, const at::Tensor & count, at::Tensor & out);
+TORCH_API at::Tensor batch_norm_backward_elemt_cuda(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & invstd, const ::std::optional<at::Tensor> & weight, const at::Tensor & sum_dy, const at::Tensor & sum_dy_xmu, const at::Tensor & count);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_reduce.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_reduce.h
new file mode 100644
index 0000000000000000000000000000000000000000..9f40d8770c3735d3cab9dd744557d9636b0299bd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_backward_reduce.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/batch_norm_backward_reduce_ops.h>
+
+namespace at {
+
+
+// aten::batch_norm_backward_reduce(Tensor grad_out, Tensor input, Tensor mean, Tensor invstd, Tensor? weight, bool input_g, bool weight_g, bool bias_g) -> (Tensor, Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> batch_norm_backward_reduce(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & invstd, const ::std::optional<at::Tensor> & weight, bool input_g, bool weight_g, bool bias_g) {
+    return at::_ops::batch_norm_backward_reduce::call(grad_out, input, mean, invstd, weight, input_g, weight_g, bias_g);
+}
+
+// aten::batch_norm_backward_reduce.out(Tensor grad_out, Tensor input, Tensor mean, Tensor invstd, Tensor? weight, bool input_g, bool weight_g, bool bias_g, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2, Tensor(d!) out3) -> (Tensor(a!), Tensor(b!), Tensor(c!), Tensor(d!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> batch_norm_backward_reduce_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3, const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & invstd, const ::std::optional<at::Tensor> & weight, bool input_g, bool weight_g, bool bias_g) {
+    return at::_ops::batch_norm_backward_reduce_out::call(grad_out, input, mean, invstd, weight, input_g, weight_g, bias_g, out0, out1, out2, out3);
+}
+// aten::batch_norm_backward_reduce.out(Tensor grad_out, Tensor input, Tensor mean, Tensor invstd, Tensor? weight, bool input_g, bool weight_g, bool bias_g, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2, Tensor(d!) out3) -> (Tensor(a!), Tensor(b!), Tensor(c!), Tensor(d!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> batch_norm_backward_reduce_outf(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & invstd, const ::std::optional<at::Tensor> & weight, bool input_g, bool weight_g, bool bias_g, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3) {
+    return at::_ops::batch_norm_backward_reduce_out::call(grad_out, input, mean, invstd, weight, input_g, weight_g, bias_g, out0, out1, out2, out3);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_gather_stats_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_gather_stats_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..caa4e8e496e3884f1bb19ac0b5e28ad3b6facbe3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_gather_stats_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> batch_norm_gather_stats(const at::Tensor & input, const at::Tensor & mean, const at::Tensor & invstd, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, double momentum, double eps, int64_t count);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_update_stats.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_update_stats.h
new file mode 100644
index 0000000000000000000000000000000000000000..89258a8e1efbc37fb93a11352df9e755e467ddd9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_update_stats.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/batch_norm_update_stats_ops.h>
+
+namespace at {
+
+
+// aten::batch_norm_update_stats(Tensor input, Tensor? running_mean, Tensor? running_var, float momentum) -> (Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor> batch_norm_update_stats(const at::Tensor & input, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, double momentum) {
+    return at::_ops::batch_norm_update_stats::call(input, running_mean, running_var, momentum);
+}
+
+// aten::batch_norm_update_stats.out(Tensor input, Tensor? running_mean, Tensor? running_var, float momentum, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))
+inline ::std::tuple<at::Tensor &,at::Tensor &> batch_norm_update_stats_out(at::Tensor & out0, at::Tensor & out1, const at::Tensor & input, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, double momentum) {
+    return at::_ops::batch_norm_update_stats_out::call(input, running_mean, running_var, momentum, out0, out1);
+}
+// aten::batch_norm_update_stats.out(Tensor input, Tensor? running_mean, Tensor? running_var, float momentum, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))
+inline ::std::tuple<at::Tensor &,at::Tensor &> batch_norm_update_stats_outf(const at::Tensor & input, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, double momentum, at::Tensor & out0, at::Tensor & out1) {
+    return at::_ops::batch_norm_update_stats_out::call(input, running_mean, running_var, momentum, out0, out1);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_update_stats_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_update_stats_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..2ba040068e7c16694a4890469e8b38b3025be2e9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/batch_norm_update_stats_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> batch_norm_update_stats(const at::Tensor & input, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, double momentum);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/binary_cross_entropy_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/binary_cross_entropy_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8b6844c567af4996903eeb73e44513910b4d85e9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/binary_cross_entropy_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 binary_cross_entropy(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight={}, int64_t reduction=at::Reduction::Mean);
+TORCH_API at::Tensor & binary_cross_entropy_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight={}, int64_t reduction=at::Reduction::Mean);
+TORCH_API at::Tensor & binary_cross_entropy_outf(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/binary_cross_entropy_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/binary_cross_entropy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..5d717de567db5bb28b56fdb85f6517a652cdf068
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/binary_cross_entropy_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 binary_cross_entropy_cpu(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight={}, int64_t reduction=at::Reduction::Mean);
+TORCH_API at::Tensor & binary_cross_entropy_out_cpu(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, at::Tensor & out);
+TORCH_API at::Tensor binary_cross_entropy_cuda(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight={}, int64_t reduction=at::Reduction::Mean);
+TORCH_API at::Tensor & binary_cross_entropy_out_cuda(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/binary_cross_entropy_with_logits_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/binary_cross_entropy_with_logits_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..073f88d687b9c8716ab66fad77bd8e2a253bc912
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/binary_cross_entropy_with_logits_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 binary_cross_entropy_with_logits(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight={}, const ::std::optional<at::Tensor> & pos_weight={}, int64_t reduction=at::Reduction::Mean);
+TORCH_API at::Tensor & binary_cross_entropy_with_logits_out(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & pos_weight, int64_t reduction, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift.h
new file mode 100644
index 0000000000000000000000000000000000000000..a505aba022ac761ac1fb034f668d41dc63e82839
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift.h
@@ -0,0 +1,73 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/bitwise_left_shift_ops.h>
+
+namespace at {
+
+
+// aten::bitwise_left_shift.Tensor(Tensor self, Tensor other) -> Tensor
+inline at::Tensor bitwise_left_shift(const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::bitwise_left_shift_Tensor::call(self, other);
+}
+
+// aten::bitwise_left_shift.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & bitwise_left_shift_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::bitwise_left_shift_Tensor_out::call(self, other, out);
+}
+// aten::bitwise_left_shift.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & bitwise_left_shift_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out) {
+    return at::_ops::bitwise_left_shift_Tensor_out::call(self, other, out);
+}
+
+// aten::bitwise_left_shift.Tensor_Scalar(Tensor self, Scalar other) -> Tensor
+inline at::Tensor bitwise_left_shift(const at::Tensor & self, const at::Scalar & other) {
+    return at::_ops::bitwise_left_shift_Tensor_Scalar::call(self, other);
+}
+
+// aten::bitwise_left_shift.Tensor_Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & bitwise_left_shift_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other) {
+    return at::_ops::bitwise_left_shift_Tensor_Scalar_out::call(self, other, out);
+}
+// aten::bitwise_left_shift.Tensor_Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & bitwise_left_shift_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out) {
+    return at::_ops::bitwise_left_shift_Tensor_Scalar_out::call(self, other, out);
+}
+
+// aten::bitwise_left_shift.Scalar_Tensor(Scalar self, Tensor other) -> Tensor
+inline at::Tensor bitwise_left_shift(const at::Scalar & self, const at::Tensor & other) {
+    return at::_ops::bitwise_left_shift_Scalar_Tensor::call(self, other);
+}
+
+// aten::bitwise_left_shift.Scalar_Tensor_out(Scalar self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & bitwise_left_shift_out(at::Tensor & out, const at::Scalar & self, const at::Tensor & other) {
+    return at::_ops::bitwise_left_shift_Scalar_Tensor_out::call(self, other, out);
+}
+// aten::bitwise_left_shift.Scalar_Tensor_out(Scalar self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & bitwise_left_shift_outf(const at::Scalar & self, const at::Tensor & other, at::Tensor & out) {
+    return at::_ops::bitwise_left_shift_Scalar_Tensor_out::call(self, other, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..90e67408a6354a3d2b770a04e369daaa47bdea32
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_left_shift(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & bitwise_left_shift_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & bitwise_left_shift_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & bitwise_left_shift_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor bitwise_left_shift(const at::Scalar & self, const at::Tensor & other);
+TORCH_API at::Tensor & bitwise_left_shift_out(at::Tensor & out, const at::Scalar & self, const at::Tensor & other);
+TORCH_API at::Tensor & bitwise_left_shift_outf(const at::Scalar & self, const at::Tensor & other, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6b2dbebde40e68a7c2a03348d2d1dc67fb3ac01d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 bitwise_left_shift(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & bitwise_left_shift_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & bitwise_left_shift_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & bitwise_left_shift_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ca89d75bbc54b9d26998fcf10dc7f274ace008e7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_left_shift_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 bitwise_left_shift(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & bitwise_left_shift_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & bitwise_left_shift_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & bitwise_left_shift_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_right_shift_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_right_shift_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..f9209b73bf4c2c83e4c2333b3b3652efc8f02e11
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bitwise_right_shift_native.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/bitwise_right_shift_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_bitwise_right_shift_out : public at::meta::structured_bitwise_right_shift_Tensor {
+void impl(const at::Tensor & self, const at::Tensor & other, const at::Tensor & out);
+};
+TORCH_API at::Tensor bitwise_right_shift(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & bitwise_right_shift_out(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & bitwise_right_shift_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor bitwise_right_shift(const at::Scalar & self, const at::Tensor & other);
+TORCH_API at::Tensor & bitwise_right_shift_Scalar_Tensor_out(const at::Scalar & self, const at::Tensor & other, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bmm_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bmm_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b3e542af0134cf94d3bfc685c36ef282f614fd65
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bmm_cuda_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 bmm(const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor & bmm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor & bmm_outf(const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out);
+TORCH_API at::Tensor bmm(const at::Tensor & self, const at::Tensor & mat2, at::ScalarType out_dtype);
+TORCH_API at::Tensor & bmm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat2, at::ScalarType out_dtype);
+TORCH_API at::Tensor & bmm_outf(const at::Tensor & self, const at::Tensor & mat2, at::ScalarType out_dtype, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bmm_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bmm_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..133c30b922d63ce644a46327a229201d8756acd7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bmm_native.h
@@ -0,0 +1,40 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/bmm_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_bmm_out_cpu : public at::meta::structured_bmm {
+void impl(const at::Tensor & self, const at::Tensor & mat2, const at::Tensor & out);
+};
+struct TORCH_API structured_bmm_out_cuda : public at::meta::structured_bmm {
+void impl(const at::Tensor & self, const at::Tensor & mat2, const at::Tensor & out);
+};
+TORCH_API at::Tensor bmm_nested(const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor bmm_nested_cuda(const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor bmm_sparse_cpu(const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor & bmm_out_sparse_cpu(const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out);
+TORCH_API at::Tensor bmm_sparse_cuda(const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor & bmm_out_sparse_cuda(const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out);
+TORCH_API at::Tensor & bmm_out_sparse_csr_cuda(const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out);
+TORCH_API at::Tensor _bmm_dtype_cuda(const at::Tensor & self, const at::Tensor & mat2, at::ScalarType out_dtype);
+TORCH_API at::Tensor & _bmm_out_dtype_cuda(const at::Tensor & self, const at::Tensor & mat2, at::ScalarType out_dtype, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bucketize_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bucketize_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b40af6335d5ad258d9e808278d8534035ea4265a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/bucketize_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 bucketize(const at::Tensor & self, const at::Tensor & boundaries, bool out_int32=false, bool right=false);
+TORCH_API at::Tensor & bucketize_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & boundaries, bool out_int32=false, bool right=false);
+TORCH_API at::Tensor & bucketize_outf(const at::Tensor & self, const at::Tensor & boundaries, bool out_int32, bool right, at::Tensor & out);
+TORCH_API at::Tensor bucketize(const at::Scalar & self, const at::Tensor & boundaries, bool out_int32=false, bool right=false);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..1bd92872efd726c8e9dc265c1da1d7e35af615ac
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 cat(const at::ITensorListRef & tensors, int64_t dim=0);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c02c7e095a4b466d716b28b7e635487b02bfe6b4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 cat(const at::ITensorListRef & tensors, int64_t dim=0);
+TORCH_API at::Tensor & cat_out(at::Tensor & out, const at::ITensorListRef & tensors, int64_t dim=0);
+TORCH_API at::Tensor & cat_outf(const at::ITensorListRef & tensors, int64_t dim, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..852b4aea82296da8ffc1c57286fb7f379e2b0e6d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_meta.h
@@ -0,0 +1,119 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_cat : public at::impl::MetaBase {
+
+                template <bool DIM = false, bool VALID = false, bool ALL_CONTIGUOUS = false, bool ALL_SAME_DTYPE = false, bool ALL_SAME_SIZES_AND_STRIDE = false, bool MEMORY_FORMAT = false>
+                struct TORCH_API precompute_out {
+
+                    precompute_out<true, VALID, ALL_CONTIGUOUS, ALL_SAME_DTYPE, ALL_SAME_SIZES_AND_STRIDE, MEMORY_FORMAT> set_dim(int64_t value) {
+                        static_assert(DIM == false, "dim already set");
+                        precompute_out<true, VALID, ALL_CONTIGUOUS, ALL_SAME_DTYPE, ALL_SAME_SIZES_AND_STRIDE, MEMORY_FORMAT> ret;
+ret.dim = value;
+ret.valid = this->valid;
+ret.all_contiguous = this->all_contiguous;
+ret.all_same_dtype = this->all_same_dtype;
+ret.all_same_sizes_and_stride = this->all_same_sizes_and_stride;
+ret.memory_format = this->memory_format;
+return ret;
+                    }
+
+
+                    precompute_out<DIM, true, ALL_CONTIGUOUS, ALL_SAME_DTYPE, ALL_SAME_SIZES_AND_STRIDE, MEMORY_FORMAT> set_valid(int64_t value) {
+                        static_assert(VALID == false, "valid already set");
+                        precompute_out<DIM, true, ALL_CONTIGUOUS, ALL_SAME_DTYPE, ALL_SAME_SIZES_AND_STRIDE, MEMORY_FORMAT> ret;
+ret.dim = this->dim;
+ret.valid = value;
+ret.all_contiguous = this->all_contiguous;
+ret.all_same_dtype = this->all_same_dtype;
+ret.all_same_sizes_and_stride = this->all_same_sizes_and_stride;
+ret.memory_format = this->memory_format;
+return ret;
+                    }
+
+
+                    precompute_out<DIM, VALID, true, ALL_SAME_DTYPE, ALL_SAME_SIZES_AND_STRIDE, MEMORY_FORMAT> set_all_contiguous(bool value) {
+                        static_assert(ALL_CONTIGUOUS == false, "all_contiguous already set");
+                        precompute_out<DIM, VALID, true, ALL_SAME_DTYPE, ALL_SAME_SIZES_AND_STRIDE, MEMORY_FORMAT> ret;
+ret.dim = this->dim;
+ret.valid = this->valid;
+ret.all_contiguous = value;
+ret.all_same_dtype = this->all_same_dtype;
+ret.all_same_sizes_and_stride = this->all_same_sizes_and_stride;
+ret.memory_format = this->memory_format;
+return ret;
+                    }
+
+
+                    precompute_out<DIM, VALID, ALL_CONTIGUOUS, true, ALL_SAME_SIZES_AND_STRIDE, MEMORY_FORMAT> set_all_same_dtype(bool value) {
+                        static_assert(ALL_SAME_DTYPE == false, "all_same_dtype already set");
+                        precompute_out<DIM, VALID, ALL_CONTIGUOUS, true, ALL_SAME_SIZES_AND_STRIDE, MEMORY_FORMAT> ret;
+ret.dim = this->dim;
+ret.valid = this->valid;
+ret.all_contiguous = this->all_contiguous;
+ret.all_same_dtype = value;
+ret.all_same_sizes_and_stride = this->all_same_sizes_and_stride;
+ret.memory_format = this->memory_format;
+return ret;
+                    }
+
+
+                    precompute_out<DIM, VALID, ALL_CONTIGUOUS, ALL_SAME_DTYPE, true, MEMORY_FORMAT> set_all_same_sizes_and_stride(bool value) {
+                        static_assert(ALL_SAME_SIZES_AND_STRIDE == false, "all_same_sizes_and_stride already set");
+                        precompute_out<DIM, VALID, ALL_CONTIGUOUS, ALL_SAME_DTYPE, true, MEMORY_FORMAT> ret;
+ret.dim = this->dim;
+ret.valid = this->valid;
+ret.all_contiguous = this->all_contiguous;
+ret.all_same_dtype = this->all_same_dtype;
+ret.all_same_sizes_and_stride = value;
+ret.memory_format = this->memory_format;
+return ret;
+                    }
+
+
+                    precompute_out<DIM, VALID, ALL_CONTIGUOUS, ALL_SAME_DTYPE, ALL_SAME_SIZES_AND_STRIDE, true> set_memory_format(at::MemoryFormat value) {
+                        static_assert(MEMORY_FORMAT == false, "memory_format already set");
+                        precompute_out<DIM, VALID, ALL_CONTIGUOUS, ALL_SAME_DTYPE, ALL_SAME_SIZES_AND_STRIDE, true> ret;
+ret.dim = this->dim;
+ret.valid = this->valid;
+ret.all_contiguous = this->all_contiguous;
+ret.all_same_dtype = this->all_same_dtype;
+ret.all_same_sizes_and_stride = this->all_same_sizes_and_stride;
+ret.memory_format = value;
+return ret;
+                    }
+
+                    int64_t dim;
+int64_t valid;
+bool all_contiguous;
+bool all_same_dtype;
+bool all_same_sizes_and_stride;
+at::MemoryFormat memory_format;
+            };
+    using meta_return_ty = precompute_out <true, true, true, true, true, true>;
+    meta_return_ty meta(const at::ITensorListRef & tensors, int64_t dim);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..0e9654a053597ed0c3de970686b2339399fb9319
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cat_native.h
@@ -0,0 +1,37 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/cat_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_cat_out_cpu : public at::meta::structured_cat {
+void impl(const at::ITensorListRef & tensors, int64_t dim, int64_t valid, bool all_contiguous, bool all_same_dtype, bool all_same_sizes_and_stride, at::MemoryFormat memory_format, const at::Tensor & out);
+};
+struct TORCH_API structured_cat_out_cuda : public at::meta::structured_cat {
+void impl(const at::ITensorListRef & tensors, int64_t dim, int64_t valid, bool all_contiguous, bool all_same_dtype, bool all_same_sizes_and_stride, at::MemoryFormat memory_format, const at::Tensor & out);
+};
+TORCH_API at::Tensor cat_nested(const at::ITensorListRef & tensors, int64_t dim=0);
+TORCH_API at::Tensor cat_sparse(const at::ITensorListRef & tensors, int64_t dim=0);
+TORCH_API at::Tensor cat_quantized_cpu(const at::ITensorListRef & tensors, int64_t dim=0);
+TORCH_API at::Tensor & cat_out_quantized_cpu(const at::ITensorListRef & tensors, int64_t dim, at::Tensor & out);
+TORCH_API at::Tensor cat(at::TensorList tensors, at::Dimname dim);
+TORCH_API at::Tensor & cat_out(at::TensorList tensors, at::Dimname dim, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cauchy_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cauchy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b6a39ff4c0a13fef71f1d1fb16419fb29e4a57da
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cauchy_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 cauchy(const at::Tensor & self, double median=0, double sigma=1, ::std::optional<at::Generator> generator=::std::nullopt);
+TORCH_API at::Tensor & cauchy_out(const at::Tensor & self, double median, double sigma, ::std::optional<at::Generator> generator, at::Tensor & out);
+TORCH_API at::Tensor & cauchy_(at::Tensor & self, double median=0, double sigma=1, ::std::optional<at::Generator> generator=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ccol_indices_copy_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ccol_indices_copy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..0b6f37dbb197272fbb1c22f4d339c9234fdd8d83
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ccol_indices_copy_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & ccol_indices_copy_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor ccol_indices_copy(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ccol_indices_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ccol_indices_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..fee4463fc2046666c59f487f69c468ba0c252f31
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ccol_indices_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API ccol_indices {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::ccol_indices";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "ccol_indices(Tensor(a) self) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ceil_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ceil_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..00af445201c69ce926af609699d945faabd2d552
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ceil_native.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/ceil_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_ceil_out : public at::meta::structured_ceil {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+TORCH_API at::Tensor ceil_sparse(const at::Tensor & self);
+TORCH_API at::Tensor & ceil_sparse_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & ceil_sparse_(at::Tensor & self);
+TORCH_API at::Tensor ceil_sparse_csr(const at::Tensor & self);
+TORCH_API at::Tensor & ceil_sparse_csr_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & ceil_sparse_csr_(at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/chalf_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/chalf_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..c672e2cdb722d6cb8a6fb9455ebdc00910f6ec2e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/chalf_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 chalf(const at::Tensor & self, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/channel_shuffle_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/channel_shuffle_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..b388ab135cd02e054ed9282571d241c0a9e21fbf
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/channel_shuffle_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 channel_shuffle {
+  using schema = at::Tensor (const at::Tensor &, c10::SymInt);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::channel_shuffle";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "channel_shuffle(Tensor self, SymInt groups) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymInt groups);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt groups);
+};
+
+struct TORCH_API channel_shuffle_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymInt, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::channel_shuffle";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "channel_shuffle.out(Tensor self, SymInt groups, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymInt groups, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt groups, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/clamp_max_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/clamp_max_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..d2148edafa6f2bbe03cae1f8cbf10eb9e5805fef
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/clamp_max_native.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/clamp_max_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_clamp_max_out : public at::meta::structured_clamp_max {
+void impl(const at::Tensor & self, const at::Scalar & max, const at::Tensor & out);
+};
+struct TORCH_API structured_clamp_max_Tensor_out : public at::meta::structured_clamp_max_Tensor {
+void impl(const at::Tensor & self, const at::Tensor & max, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/clamp_min_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/clamp_min_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d661d29f41150246ca0d50517c37a5789aca5b8a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/clamp_min_meta_dispatch.h
@@ -0,0 +1,35 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API at::Tensor clamp_min(const at::Tensor & self, const at::Scalar & min);
+TORCH_API at::Tensor & clamp_min_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & min);
+TORCH_API at::Tensor & clamp_min_outf(const at::Tensor & self, const at::Scalar & min, at::Tensor & out);
+TORCH_API at::Tensor & clamp_min_(at::Tensor & self, const at::Scalar & min);
+TORCH_API at::Tensor clamp_min(const at::Tensor & self, const at::Tensor & min);
+TORCH_API at::Tensor & clamp_min_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & min);
+TORCH_API at::Tensor & clamp_min_outf(const at::Tensor & self, const at::Tensor & min, at::Tensor & out);
+TORCH_API at::Tensor & clamp_min_(at::Tensor & self, const at::Tensor & min);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/clone.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/clone.h
new file mode 100644
index 0000000000000000000000000000000000000000..51da74339a6c3f14c9e5da0f1bbcfd3ec26fcbd0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/clone.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/clone_ops.h>
+
+namespace at {
+
+
+// aten::clone(Tensor self, *, MemoryFormat? memory_format=None) -> Tensor
+inline at::Tensor clone(const at::Tensor & self, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt) {
+    return at::_ops::clone::call(self, memory_format);
+}
+
+// aten::clone.out(Tensor self, *, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & clone_out(at::Tensor & out, const at::Tensor & self, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt) {
+    return at::_ops::clone_out::call(self, memory_format, out);
+}
+// aten::clone.out(Tensor self, *, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & clone_outf(const at::Tensor & self, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out) {
+    return at::_ops::clone_out::call(self, memory_format, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/col2im_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/col2im_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c0bd850942970317d7a5925ee38883c26eb59337
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/col2im_cuda_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 col2im(const at::Tensor & self, at::IntArrayRef output_size, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor col2im_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor & col2im_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor & col2im_outf(const at::Tensor & self, at::IntArrayRef output_size, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride, at::Tensor & out);
+TORCH_API at::Tensor & col2im_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor & col2im_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/column_stack.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/column_stack.h
new file mode 100644
index 0000000000000000000000000000000000000000..a0e7ed4b15eb97770b42b482058f03e741b31577
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/column_stack.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/column_stack_ops.h>
+
+namespace at {
+
+
+// aten::column_stack(Tensor[] tensors) -> Tensor
+inline at::Tensor column_stack(at::TensorList tensors) {
+    return at::_ops::column_stack::call(tensors);
+}
+
+// aten::column_stack.out(Tensor[] tensors, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & column_stack_out(at::Tensor & out, at::TensorList tensors) {
+    return at::_ops::column_stack_out::call(tensors, out);
+}
+// aten::column_stack.out(Tensor[] tensors, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & column_stack_outf(at::TensorList tensors, at::Tensor & out) {
+    return at::_ops::column_stack_out::call(tensors, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/column_stack_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/column_stack_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..fa36e66d1ff29e93173975a0c4ca8a7ea3d8c754
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/column_stack_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 column_stack {
+  using schema = at::Tensor (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::column_stack";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "column_stack(Tensor[] tensors) -> Tensor";
+  static at::Tensor call(at::TensorList tensors);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList tensors);
+};
+
+struct TORCH_API column_stack_out {
+  using schema = at::Tensor & (at::TensorList, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::column_stack";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "column_stack.out(Tensor[] tensors, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(at::TensorList tensors, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList tensors, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/concat.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/concat.h
new file mode 100644
index 0000000000000000000000000000000000000000..935a99d69434c7165240f150fefd95cf08baddeb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/concat.h
@@ -0,0 +1,59 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/concat_ops.h>
+
+namespace at {
+
+
+// aten::concat(Tensor[] tensors, int dim=0) -> Tensor
+inline at::Tensor concat(at::TensorList tensors, int64_t dim=0) {
+    return at::_ops::concat::call(tensors, dim);
+}
+
+// aten::concat.out(Tensor[] tensors, int dim=0, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & concat_out(at::Tensor & out, at::TensorList tensors, int64_t dim=0) {
+    return at::_ops::concat_out::call(tensors, dim, out);
+}
+// aten::concat.out(Tensor[] tensors, int dim=0, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & concat_outf(at::TensorList tensors, int64_t dim, at::Tensor & out) {
+    return at::_ops::concat_out::call(tensors, dim, out);
+}
+
+// aten::concat.names(Tensor[] tensors, Dimname dim) -> Tensor
+inline at::Tensor concat(at::TensorList tensors, at::Dimname dim) {
+    return at::_ops::concat_names::call(tensors, dim);
+}
+
+// aten::concat.names_out(Tensor[] tensors, Dimname dim, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & concat_out(at::Tensor & out, at::TensorList tensors, at::Dimname dim) {
+    return at::_ops::concat_names_out::call(tensors, dim, out);
+}
+// aten::concat.names_out(Tensor[] tensors, Dimname dim, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & concat_outf(at::TensorList tensors, at::Dimname dim, at::Tensor & out) {
+    return at::_ops::concat_names_out::call(tensors, dim, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/constant_pad_nd_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/constant_pad_nd_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..5d157e214810c414fecf09265a055f41580deb5b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/constant_pad_nd_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 constant_pad_nd(const at::Tensor & self, at::IntArrayRef pad, const at::Scalar & value=0);
+TORCH_API at::Tensor & constant_pad_nd_out_symint(const at::Tensor & self, c10::SymIntArrayRef pad, const at::Scalar & value, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/contiguous_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/contiguous_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..34f0f5a748cdbcb99091ec6e42595edd298dcba9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/contiguous_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 contiguous(const at::Tensor & self, at::MemoryFormat memory_format=c10::MemoryFormat::Contiguous);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/contiguous_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/contiguous_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..ca21276fb318b1e70644ecfb5aaf1e06261cae1e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/contiguous_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 contiguous {
+  using schema = at::Tensor (const at::Tensor &, at::MemoryFormat);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::contiguous";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "contiguous(Tensor(a) self, *, MemoryFormat memory_format=contiguous_format) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, at::MemoryFormat memory_format);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::MemoryFormat memory_format);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/conv3d_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/conv3d_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..43ba8a241332678d846f29e95a19a41bbf7c4f6d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/conv3d_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 conv3d_symint(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias={}, c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef dilation=c10::SymInt(1), c10::SymInt groups=1);
+TORCH_API at::Tensor conv3d_padding_symint(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias={}, c10::SymIntArrayRef stride=c10::SymInt(1), c10::string_view padding="valid", c10::SymIntArrayRef dilation=c10::SymInt(1), c10::SymInt groups=1);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/conv3d_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/conv3d_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..a81e07e53abc75175d1b337d9d634648a829cf46
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/conv3d_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 conv3d {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, 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 const char* name = "aten::conv3d";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "conv3d(Tensor input, Tensor weight, Tensor? bias=None, SymInt[3] stride=1, SymInt[3] padding=0, SymInt[3] dilation=1, SymInt groups=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & input, const at::Tensor & weight, 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 & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, c10::SymInt groups);
+};
+
+struct TORCH_API conv3d_padding {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::string_view, c10::SymIntArrayRef, c10::SymInt);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::conv3d";
+  static constexpr const char* overload_name = "padding";
+  static constexpr const char* schema_str = "conv3d.padding(Tensor input, Tensor weight, Tensor? bias=None, SymInt[3] stride=1, str padding=\"valid\", SymInt[3] dilation=1, SymInt groups=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::string_view padding, c10::SymIntArrayRef dilation, c10::SymInt groups);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::string_view padding, c10::SymIntArrayRef dilation, c10::SymInt groups);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/conv_transpose1d_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/conv_transpose1d_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..92b1eb4a4cfebd080de1d33f7f9b39645c13fbad
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/conv_transpose1d_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 conv_transpose1d_symint(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias={}, c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef output_padding=c10::SymInt(0), c10::SymInt groups=1, c10::SymIntArrayRef dilation=c10::SymInt(1));
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f071c35a88475dffbe1224d73147a8ad2282eb50
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_cuda_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> convolution_backward(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, at::OptionalIntArrayRef bias_sizes, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool transposed, at::IntArrayRef output_padding, int64_t groups, ::std::array<bool,3> output_mask);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> convolution_backward_symint(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, at::OptionalSymIntArrayRef bias_sizes, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..e478de7cdc6a18b586d014a1ee5f38e50f336959
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> convolution_backward(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, at::OptionalIntArrayRef bias_sizes, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool transposed, at::IntArrayRef output_padding, int64_t groups, ::std::array<bool,3> output_mask);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> convolution_backward_out_symint(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, at::OptionalSymIntArrayRef bias_sizes, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_overrideable_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_overrideable_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5b2d0ab23a703e2c2dcca849fc005b1e4ce49a3c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_overrideable_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> convolution_backward_overrideable(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool transposed, at::IntArrayRef output_padding, int64_t groups, ::std::array<bool,3> output_mask);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> convolution_backward_overrideable_symint(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> convolution_backward_overrideable_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool transposed, at::IntArrayRef output_padding, int64_t groups, ::std::array<bool,3> output_mask);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> convolution_backward_overrideable_outf(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool transposed, at::IntArrayRef output_padding, int64_t groups, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> convolution_backward_overrideable_symint_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> convolution_backward_overrideable_symint_outf(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_overrideable_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_overrideable_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..0d9911fdb0ddb59c6ded4efdcb21c050449d2d4e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_backward_overrideable_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 convolution_backward_overrideable {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, c10::SymIntArrayRef, c10::SymInt, ::std::array<bool,3>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::convolution_backward_overrideable";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "convolution_backward_overrideable(Tensor grad_output, Tensor input, Tensor weight, SymInt[] stride, SymInt[] padding, SymInt[] dilation, bool transposed, SymInt[] output_padding, SymInt groups, bool[3] output_mask) -> (Tensor grad_input, Tensor grad_weight, Tensor grad_bias)";
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> call(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask);
+};
+
+struct TORCH_API convolution_backward_overrideable_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> (const at::Tensor &, const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, c10::SymIntArrayRef, c10::SymInt, ::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 const char* name = "aten::convolution_backward_overrideable";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "convolution_backward_overrideable.out(Tensor grad_output, Tensor input, Tensor weight, SymInt[] stride, SymInt[] padding, SymInt[] dilation, bool transposed, SymInt[] output_padding, SymInt groups, 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_output, const at::Tensor & input, const at::Tensor & weight, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::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_output, const at::Tensor & input, const at::Tensor & weight, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..134b936ef05a0cb9abd6339723d70ac144f08e3c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 convolution {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, c10::SymIntArrayRef, c10::SymInt);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::convolution";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "convolution(Tensor input, Tensor weight, Tensor? bias, SymInt[] stride, SymInt[] padding, SymInt[] dilation, bool transposed, SymInt[] output_padding, SymInt groups) -> Tensor";
+  static at::Tensor call(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups);
+};
+
+struct TORCH_API convolution_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, c10::SymIntArrayRef, c10::SymInt, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::convolution";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "convolution.out(Tensor input, Tensor weight, Tensor? bias, SymInt[] stride, SymInt[] padding, SymInt[] dilation, bool transposed, SymInt[] output_padding, SymInt groups, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_overrideable_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_overrideable_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..bfc5e2974e966e30e2df127e56caffd4cfd41ab2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/convolution_overrideable_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 convolution_overrideable(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool transposed, at::IntArrayRef output_padding, int64_t groups);
+TORCH_API at::Tensor & convolution_overrideable_out_symint(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/copy_sparse_to_sparse_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/copy_sparse_to_sparse_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..cfd214b12842861214ed65c612460ab458e5ac5a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/copy_sparse_to_sparse_meta_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & copy_sparse_to_sparse_(at::Tensor & self, const at::Tensor & src, bool non_blocking=false);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/copysign.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/copysign.h
new file mode 100644
index 0000000000000000000000000000000000000000..263c3a531525c24cf6026a55d65e1368208a6459
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/copysign.h
@@ -0,0 +1,59 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/copysign_ops.h>
+
+namespace at {
+
+
+// aten::copysign.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & copysign_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::copysign_out::call(self, other, out);
+}
+// aten::copysign.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & copysign_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out) {
+    return at::_ops::copysign_out::call(self, other, out);
+}
+
+// aten::copysign.Tensor(Tensor self, Tensor other) -> Tensor
+inline at::Tensor copysign(const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::copysign_Tensor::call(self, other);
+}
+
+// aten::copysign.Scalar(Tensor self, Scalar other) -> Tensor
+inline at::Tensor copysign(const at::Tensor & self, const at::Scalar & other) {
+    return at::_ops::copysign_Scalar::call(self, other);
+}
+
+// aten::copysign.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & copysign_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other) {
+    return at::_ops::copysign_Scalar_out::call(self, other, out);
+}
+// aten::copysign.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & copysign_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out) {
+    return at::_ops::copysign_Scalar_out::call(self, other, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cos_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cos_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..1dd5287f53c8cbc982b43af32826bb542cce30a7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cos_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/cos_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_cos_out : public at::meta::structured_cos {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+TORCH_API at::Tensor NestedTensor_cos(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cosh_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cosh_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..37a962fd77a064e9584857d9f82e9a748b90e38b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cosh_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 cosh(const at::Tensor & self);
+TORCH_API at::Tensor & cosh_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cosh_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cosh_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4fe5bd02e9544656b52595f8aeab10b45f42286d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cosh_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 cosh(const at::Tensor & self);
+TORCH_API at::Tensor & cosh_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & cosh_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & cosh_(at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cov.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cov.h
new file mode 100644
index 0000000000000000000000000000000000000000..bf109864c17de3123a1126a749d918e6c4eb949e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cov.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/cov_ops.h>
+
+namespace at {
+
+
+// aten::cov(Tensor self, *, int correction=1, Tensor? fweights=None, Tensor? aweights=None) -> Tensor
+inline at::Tensor cov(const at::Tensor & self, int64_t correction=1, const ::std::optional<at::Tensor> & fweights={}, const ::std::optional<at::Tensor> & aweights={}) {
+    return at::_ops::cov::call(self, correction, fweights, aweights);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cross_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cross_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..694032c09d49a40ca1598da8751d932867d2d35f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cross_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 cross(const at::Tensor & self, const at::Tensor & other, ::std::optional<int64_t> dim=::std::nullopt);
+TORCH_API at::Tensor & cross_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other, ::std::optional<int64_t> dim=::std::nullopt);
+TORCH_API at::Tensor & cross_outf(const at::Tensor & self, const at::Tensor & other, ::std::optional<int64_t> dim, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cross_entropy_loss_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cross_entropy_loss_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..7d263efbe8abeab44601a5f997157ede7c64b8a1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cross_entropy_loss_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 cross_entropy_loss {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, int64_t, c10::SymInt, double);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::cross_entropy_loss";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "cross_entropy_loss(Tensor self, Tensor target, Tensor? weight=None, int reduction=Mean, SymInt ignore_index=-100, float label_smoothing=0.0) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index, double label_smoothing);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index, double label_smoothing);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/crow_indices.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/crow_indices.h
new file mode 100644
index 0000000000000000000000000000000000000000..807162ef475fe5f91c30c624d51af68fda23dac4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/crow_indices.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/crow_indices_ops.h>
+
+namespace at {
+
+
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/crow_indices_copy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/crow_indices_copy.h
new file mode 100644
index 0000000000000000000000000000000000000000..992ba05f3bd3489ffb76a75e33f4a19514c4c08a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/crow_indices_copy.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/crow_indices_copy_ops.h>
+
+namespace at {
+
+
+// aten::crow_indices_copy(Tensor self) -> Tensor
+inline at::Tensor crow_indices_copy(const at::Tensor & self) {
+    return at::_ops::crow_indices_copy::call(self);
+}
+
+// aten::crow_indices_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & crow_indices_copy_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::crow_indices_copy_out::call(self, out);
+}
+// aten::crow_indices_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & crow_indices_copy_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::crow_indices_copy_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_batch_norm_backward_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_batch_norm_backward_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f3105428ce9875d1f4d7edfc7ac35bfac29d461e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_batch_norm_backward_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> cudnn_batch_norm_backward_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & input, const at::Tensor & grad_output, const 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_var, double epsilon, const at::Tensor & reserveSpace);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> cudnn_batch_norm_backward_outf(const at::Tensor & input, const at::Tensor & grad_output, const 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_var, double epsilon, const at::Tensor & reserveSpace, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_batch_norm_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_batch_norm_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b37d9b85eb51c213b84fb3530c694daf5230708d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_batch_norm_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> cudnn_batch_norm(const at::Tensor & input, const 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 exponential_average_factor, double epsilon);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> cudnn_batch_norm_out(const at::Tensor & input, const 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 exponential_average_factor, double epsilon, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution.h
new file mode 100644
index 0000000000000000000000000000000000000000..013b50ee22b715d60074bf726ae58a26f9dda43b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/cudnn_convolution_ops.h>
+
+namespace at {
+
+
+// aten::cudnn_convolution(Tensor self, Tensor weight, SymInt[] padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic, bool allow_tf32) -> Tensor
+inline at::Tensor cudnn_convolution(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32) {
+    return at::_ops::cudnn_convolution::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, benchmark, deterministic, allow_tf32);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor cudnn_convolution(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32) {
+    return at::_ops::cudnn_convolution::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, benchmark, deterministic, allow_tf32);
+  }
+}
+
+// aten::cudnn_convolution(Tensor self, Tensor weight, SymInt[] padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic, bool allow_tf32) -> Tensor
+inline at::Tensor cudnn_convolution_symint(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32) {
+    return at::_ops::cudnn_convolution::call(self, weight, padding, stride, dilation, groups, benchmark, deterministic, allow_tf32);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor cudnn_convolution(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32) {
+    return at::_ops::cudnn_convolution::call(self, weight, padding, stride, dilation, groups, benchmark, deterministic, allow_tf32);
+  }
+}
+
+// aten::cudnn_convolution.out(Tensor self, Tensor weight, SymInt[] padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & cudnn_convolution_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32) {
+    return at::_ops::cudnn_convolution_out::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, benchmark, deterministic, allow_tf32, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & cudnn_convolution_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32) {
+    return at::_ops::cudnn_convolution_out::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, benchmark, deterministic, allow_tf32, out);
+  }
+}
+
+// aten::cudnn_convolution.out(Tensor self, Tensor weight, SymInt[] padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & cudnn_convolution_outf(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out) {
+    return at::_ops::cudnn_convolution_out::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, benchmark, deterministic, allow_tf32, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & cudnn_convolution_outf(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out) {
+    return at::_ops::cudnn_convolution_out::call(self, weight, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, benchmark, deterministic, allow_tf32, out);
+  }
+}
+
+// aten::cudnn_convolution.out(Tensor self, Tensor weight, SymInt[] padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & cudnn_convolution_symint_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32) {
+    return at::_ops::cudnn_convolution_out::call(self, weight, padding, stride, dilation, groups, benchmark, deterministic, allow_tf32, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & cudnn_convolution_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32) {
+    return at::_ops::cudnn_convolution_out::call(self, weight, padding, stride, dilation, groups, benchmark, deterministic, allow_tf32, out);
+  }
+}
+
+// aten::cudnn_convolution.out(Tensor self, Tensor weight, SymInt[] padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & cudnn_convolution_symint_outf(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out) {
+    return at::_ops::cudnn_convolution_out::call(self, weight, padding, stride, dilation, groups, benchmark, deterministic, allow_tf32, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & cudnn_convolution_outf(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out) {
+    return at::_ops::cudnn_convolution_out::call(self, weight, padding, stride, dilation, groups, benchmark, deterministic, allow_tf32, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_add_relu_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_add_relu_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..1258e8bfafc3fae919d1a46ba8b6b359e1e3351a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_add_relu_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & cudnn_convolution_add_relu_out_symint(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, at::Tensor & out);
+TORCH_API at::Tensor cudnn_convolution_add_relu(const at::Tensor & self, const at::Tensor & weight, const at::Tensor & z, const ::std::optional<at::Scalar> & alpha, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, int64_t groups);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..64cc498945307084221be294631d54fb43da206f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_cuda_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 cudnn_convolution(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32);
+TORCH_API at::Tensor cudnn_convolution_symint(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32);
+TORCH_API at::Tensor & cudnn_convolution_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32);
+TORCH_API at::Tensor & cudnn_convolution_outf(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out);
+TORCH_API at::Tensor & cudnn_convolution_symint_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32);
+TORCH_API at::Tensor & cudnn_convolution_symint_outf(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..dcf27dd0cb76da9f7b1b764e927bf495517b0f8e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 cudnn_convolution(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32);
+TORCH_API at::Tensor & cudnn_convolution_out(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..9fd044f10f426917871c52a76fff3ea940d5f8e6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 cudnn_convolution {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymInt, bool, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::cudnn_convolution";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "cudnn_convolution(Tensor self, Tensor weight, SymInt[] padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic, bool allow_tf32) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32);
+};
+
+struct TORCH_API cudnn_convolution_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymInt, bool, bool, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::cudnn_convolution";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "cudnn_convolution.out(Tensor self, Tensor weight, SymInt[] padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_transpose_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_transpose_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..8ab3a366434752415613e978c8bb63c038be7388
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_transpose_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & cudnn_convolution_transpose_out_symint(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out);
+TORCH_API at::Tensor cudnn_convolution_transpose(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef padding, at::IntArrayRef output_padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic, bool allow_tf32);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_transpose_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_transpose_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..1edd5a8a89ec70e7c3adfdfa1f740abbc8123464
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cudnn_convolution_transpose_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 cudnn_convolution_transpose {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymInt, bool, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::cudnn_convolution_transpose";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "cudnn_convolution_transpose(Tensor self, Tensor weight, SymInt[] padding, SymInt[] output_padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic, bool allow_tf32) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32);
+};
+
+struct TORCH_API cudnn_convolution_transpose_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymInt, bool, bool, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::cudnn_convolution_transpose";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "cudnn_convolution_transpose.out(Tensor self, Tensor weight, SymInt[] padding, SymInt[] output_padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic, bool allow_tf32, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cummaxmin_backward_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cummaxmin_backward_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..cbc1d3634ac7a6d8396177d2e9ca15966af2aa1a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cummaxmin_backward_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 cummaxmin_backward(const at::Tensor & grad, const at::Tensor & input, const at::Tensor & indices, int64_t dim);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e7a86e18c501674dc1dea19e31b4c06cdde7e0ba
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 cumprod(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & cumprod_(at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b8edd9fa6500d41801fb8bd5b40b5cc800e926ab
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 cumprod(const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & cumprod_out(at::Tensor & out, const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & cumprod_outf(const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+TORCH_API at::Tensor & cumprod_(at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..7f60c3112a463fc817b0669f2745bc9c4b03059e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor cumprod(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & cumprod_out(at::Tensor & out, const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & cumprod_outf(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+TORCH_API at::Tensor & cumprod_(at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..f953894baab92cd6a147dc70a98bc0f87b994ba7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumprod_native.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/cumprod_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_cumprod_out : public at::meta::structured_cumprod {
+void impl(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype, const at::Tensor & out);
+};
+TORCH_API at::Tensor cumprod(const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & cumprod_out(const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+TORCH_API at::Tensor & cumprod_(at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumsum.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumsum.h
new file mode 100644
index 0000000000000000000000000000000000000000..b7c5530298807c987546cd73a3580889a22fb896
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumsum.h
@@ -0,0 +1,59 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/cumsum_ops.h>
+
+namespace at {
+
+
+// aten::cumsum(Tensor self, int dim, *, ScalarType? dtype=None) -> Tensor
+inline at::Tensor cumsum(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt) {
+    return at::_ops::cumsum::call(self, dim, dtype);
+}
+
+// aten::cumsum.out(Tensor self, int dim, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & cumsum_out(at::Tensor & out, const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt) {
+    return at::_ops::cumsum_out::call(self, dim, dtype, out);
+}
+// aten::cumsum.out(Tensor self, int dim, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & cumsum_outf(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype, at::Tensor & out) {
+    return at::_ops::cumsum_out::call(self, dim, dtype, out);
+}
+
+// aten::cumsum.dimname(Tensor self, Dimname dim, *, ScalarType? dtype=None) -> Tensor
+inline at::Tensor cumsum(const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype=::std::nullopt) {
+    return at::_ops::cumsum_dimname::call(self, dim, dtype);
+}
+
+// aten::cumsum.dimname_out(Tensor self, Dimname dim, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & cumsum_out(at::Tensor & out, const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype=::std::nullopt) {
+    return at::_ops::cumsum_dimname_out::call(self, dim, dtype, out);
+}
+// aten::cumsum.dimname_out(Tensor self, Dimname dim, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & cumsum_outf(const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype, at::Tensor & out) {
+    return at::_ops::cumsum_dimname_out::call(self, dim, dtype, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumsum_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumsum_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..d0ba0e7caf87c801cff6246e28f2ea13c6a30e6a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/cumsum_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_cumsum : public at::impl::MetaBase {
+
+
+    void meta(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/data_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/data_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..f0946c775e6c9354a2c37e8fc7ab31c74a8a2a45
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/data_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 data(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/deg2rad_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/deg2rad_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6beba2593ea090940097edc099973de86f429354
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/deg2rad_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 deg2rad(const at::Tensor & self);
+TORCH_API at::Tensor & deg2rad_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & deg2rad_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & deg2rad_(at::Tensor & self);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/dequantize_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/dequantize_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..1fda5ebe6fa3dc5c48a9d5e3e7c6fe32c9ed1b89
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/dequantize_ops.h
@@ -0,0 +1,67 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 dequantize_self {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::dequantize";
+  static constexpr const char* overload_name = "self";
+  static constexpr const char* schema_str = "dequantize.self(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 dequantize_tensors {
+  using schema = ::std::vector<at::Tensor> (at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::dequantize";
+  static constexpr const char* overload_name = "tensors";
+  static constexpr const char* schema_str = "dequantize.tensors(Tensor[] tensors) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(at::TensorList tensors);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList tensors);
+};
+
+struct TORCH_API dequantize_self_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 const char* name = "aten::dequantize";
+  static constexpr const char* overload_name = "self_out";
+  static constexpr const char* schema_str = "dequantize.self_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);
+};
+
+struct TORCH_API dequantize_tensors_out {
+  using schema = void (at::TensorList, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::dequantize";
+  static constexpr const char* overload_name = "tensors_out";
+  static constexpr const char* schema_str = "dequantize.tensors_out(Tensor[] tensors, *, Tensor(a!)[] out) -> ()";
+  static void call(at::TensorList tensors, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList tensors, at::TensorList out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/det_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/det_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..420c87de51cd925b6a3682d9fd0c42ef76d22fd6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/det_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 det(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/det_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/det_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..d7713245056e10c21701b9a4febf6065ccd0575c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/det_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 det {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::det";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "det(Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/detach_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/detach_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..36eab776fe3f185b2931a8752eae54bd0ba57b8b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/detach_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 detach(const at::Tensor & self);
+TORCH_API at::Tensor & detach_(at::Tensor & self);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/detach_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/detach_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..8f619cbb4ac9a5ac11d63df09d500229ff422313
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/detach_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 detach(const at::Tensor & self);
+TORCH_API at::Tensor & detach_(at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/diag_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/diag_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..1c65b12688d22aa581b371f19d063543e35fa096
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/diag_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 diag_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::diag";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "diag.out(Tensor self, int diagonal=0, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, int64_t diagonal, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t diagonal, at::Tensor & out);
+};
+
+struct TORCH_API diag {
+  using schema = at::Tensor (const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::diag";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "diag(Tensor self, int diagonal=0) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t diagonal);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t diagonal);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/diagonal_scatter_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/diagonal_scatter_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..53dbeb613a55011bdfd75f8f14532af2b54dfb8a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/diagonal_scatter_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 diagonal_scatter {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, int64_t, int64_t, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::diagonal_scatter";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "diagonal_scatter(Tensor self, Tensor src, int offset=0, int dim1=0, int dim2=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & src, int64_t offset, int64_t dim1, int64_t dim2);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & src, int64_t offset, int64_t dim1, int64_t dim2);
+};
+
+struct TORCH_API diagonal_scatter_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, int64_t, int64_t, int64_t, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::diagonal_scatter";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "diagonal_scatter.out(Tensor self, Tensor src, int offset=0, int dim1=0, int dim2=1, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & src, int64_t offset, int64_t dim1, int64_t dim2, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & src, int64_t offset, int64_t dim1, int64_t dim2, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/diff_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/diff_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..58ab4147b86feaac8f6f5af212545d5b0cc8472d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/diff_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 diff(const at::Tensor & self, int64_t n=1, int64_t dim=-1, const ::std::optional<at::Tensor> & prepend={}, const ::std::optional<at::Tensor> & append={});
+TORCH_API at::Tensor & diff_out(at::Tensor & out, const at::Tensor & self, int64_t n=1, int64_t dim=-1, const ::std::optional<at::Tensor> & prepend={}, const ::std::optional<at::Tensor> & append={});
+TORCH_API at::Tensor & diff_outf(const at::Tensor & self, int64_t n, int64_t dim, const ::std::optional<at::Tensor> & prepend, const ::std::optional<at::Tensor> & append, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/div_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/div_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..144dc620cb7d2121a4fa09f71b349906c4194706
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/div_cpu_dispatch.h
@@ -0,0 +1,35 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 div(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & div_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & div_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & div_(at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor div(const at::Tensor & self, const at::Tensor & other, ::std::optional<c10::string_view> rounding_mode);
+TORCH_API at::Tensor & div_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other, ::std::optional<c10::string_view> rounding_mode);
+TORCH_API at::Tensor & div_outf(const at::Tensor & self, const at::Tensor & other, ::std::optional<c10::string_view> rounding_mode, at::Tensor & out);
+TORCH_API at::Tensor & div_(at::Tensor & self, const at::Tensor & other, ::std::optional<c10::string_view> rounding_mode);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/dropout_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/dropout_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..fbf3f4548e702ccd80aabfb7615300574b45b20d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/dropout_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 dropout {
+  using schema = at::Tensor (const at::Tensor &, double, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::dropout";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "dropout(Tensor input, float p, bool train) -> Tensor";
+  static at::Tensor call(const at::Tensor & input, double p, bool train);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, double p, bool train);
+};
+
+struct TORCH_API dropout_ {
+  using schema = at::Tensor & (at::Tensor &, double, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::dropout_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "dropout_(Tensor(a!) self, float p, bool train) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, double p, bool train);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, double p, bool train);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/elu_backward_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/elu_backward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4e18f157ca2b0f677863754c885052c16bdc4c34
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/elu_backward_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 elu_backward(const at::Tensor & grad_output, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale, bool is_result, const at::Tensor & self_or_result);
+TORCH_API at::Tensor & elu_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale, bool is_result, const at::Tensor & self_or_result);
+TORCH_API at::Tensor & elu_backward_outf(const at::Tensor & grad_output, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale, bool is_result, const at::Tensor & self_or_result, at::Tensor & grad_input);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/elu_backward_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/elu_backward_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4b8bba62a172a7fe5a382da8ba2d940353ac2300
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/elu_backward_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 elu_backward(const at::Tensor & grad_output, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale, bool is_result, const at::Tensor & self_or_result);
+TORCH_API at::Tensor & elu_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale, bool is_result, const at::Tensor & self_or_result);
+TORCH_API at::Tensor & elu_backward_outf(const at::Tensor & grad_output, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale, bool is_result, const at::Tensor & self_or_result, at::Tensor & grad_input);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/elu_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/elu_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..52fbcca9fa955b9bac050068145687588cd82195
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/elu_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 elu_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Scalar &, const at::Scalar &, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::elu";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "elu.out(Tensor self, Scalar alpha=1, Scalar scale=1, Scalar input_scale=1, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale, at::Tensor & out);
+};
+
+struct TORCH_API elu {
+  using schema = at::Tensor (const at::Tensor &, const at::Scalar &, const at::Scalar &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::elu";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "elu(Tensor self, Scalar alpha=1, Scalar scale=1, Scalar input_scale=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale);
+};
+
+struct TORCH_API elu_ {
+  using schema = at::Tensor & (at::Tensor &, const at::Scalar &, const at::Scalar &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::elu_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "elu_(Tensor(a!) self, Scalar alpha=1, Scalar scale=1, Scalar input_scale=1) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..1a4a30935913ec08d006465413dcbf8b36c084fd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_backward_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 embedding_backward {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, c10::SymInt, c10::SymInt, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::embedding_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "embedding_backward(Tensor grad, Tensor indices, SymInt num_weights, SymInt padding_idx, bool scale_grad_by_freq, bool sparse) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad, const at::Tensor & indices, c10::SymInt num_weights, c10::SymInt padding_idx, bool scale_grad_by_freq, bool sparse);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad, const at::Tensor & indices, c10::SymInt num_weights, c10::SymInt padding_idx, bool scale_grad_by_freq, bool sparse);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_bag.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_bag.h
new file mode 100644
index 0000000000000000000000000000000000000000..da91f3d3cebe9b3e6d5547ebe229a3bd8d48f887
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_bag.h
@@ -0,0 +1,41 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/embedding_bag_ops.h>
+
+namespace at {
+
+
+// aten::embedding_bag(Tensor weight, Tensor indices, Tensor offsets, bool scale_grad_by_freq=False, int mode=0, bool sparse=False, Tensor? per_sample_weights=None, bool include_last_offset=False) -> (Tensor, Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> embedding_bag(const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, bool scale_grad_by_freq=false, int64_t mode=0, bool sparse=false, const ::std::optional<at::Tensor> & per_sample_weights={}, bool include_last_offset=false) {
+    return at::_ops::embedding_bag::call(weight, indices, offsets, scale_grad_by_freq, mode, sparse, per_sample_weights, include_last_offset);
+}
+
+// aten::embedding_bag.padding_idx(Tensor weight, Tensor indices, Tensor offsets, bool scale_grad_by_freq, int mode, bool sparse, Tensor? per_sample_weights, bool include_last_offset, int? padding_idx) -> (Tensor, Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> embedding_bag(const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, bool scale_grad_by_freq, int64_t mode, bool sparse, const ::std::optional<at::Tensor> & per_sample_weights, bool include_last_offset, ::std::optional<int64_t> padding_idx) {
+    return at::_ops::embedding_bag_padding_idx::call(weight, indices, offsets, scale_grad_by_freq, mode, sparse, per_sample_weights, include_last_offset, padding_idx);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_dense_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_dense_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..31365ffba565b001643041003f8b5f4540154554
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_dense_backward_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_dense_backward_out_symint(const at::Tensor & grad_output, const at::Tensor & indices, c10::SymInt num_weights, c10::SymInt padding_idx, bool scale_grad_by_freq, at::Tensor & out);
+TORCH_API at::Tensor embedding_dense_backward_cpu(const at::Tensor & grad_output, const at::Tensor & indices, int64_t num_weights, int64_t padding_idx, bool scale_grad_by_freq);
+TORCH_API at::Tensor embedding_dense_backward_cuda(const at::Tensor & grad_output, const at::Tensor & indices, int64_t num_weights, int64_t padding_idx, bool scale_grad_by_freq);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_renorm_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_renorm_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..98e302cc5eb7dddb48c9d7d1d01df222d81c93f6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_renorm_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & embedding_renorm_(at::Tensor & self, const at::Tensor & indices, double max_norm, double norm_type);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_renorm_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_renorm_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..99d8177e45586cafd4b2a4271b55075b4b86c36a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_renorm_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_sparse_backward_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_sparse_backward_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..eea9baca0212e2358eeaa43ccb47be00dcfbba55
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/embedding_sparse_backward_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 embedding_sparse_backward(const at::Tensor & grad, const at::Tensor & indices, int64_t num_weights, int64_t padding_idx, bool scale_grad_by_freq);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ab0f2932d6233a6e98fb9bc799c8ab63e32b42aa
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 empty(at::IntArrayRef size, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor empty(at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+TORCH_API at::Tensor empty_symint(c10::SymIntArrayRef size, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor empty_symint(c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_quantized_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_quantized_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..59f56b8ca4a4e37718abf651ec363807a55d9038
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_quantized_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & empty_quantized_out(at::IntArrayRef size, const at::Tensor & qtensor, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor empty_quantized(at::IntArrayRef size, const at::Tensor & qtensor, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_strided_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_strided_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6d37c26e4a9842c8415dea6d9b90ab1932ed491c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_strided_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 empty_strided(at::IntArrayRef size, at::IntArrayRef stride, at::TensorOptions options={});
+TORCH_API at::Tensor empty_strided(at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+TORCH_API at::Tensor empty_strided_symint(c10::SymIntArrayRef size, c10::SymIntArrayRef stride, at::TensorOptions options={});
+TORCH_API at::Tensor empty_strided_symint(c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_strided_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_strided_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a1c1de4b1aaa29a53eeb472bf56ad20e32393fa1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/empty_strided_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 empty_strided(at::IntArrayRef size, at::IntArrayRef stride, at::TensorOptions options={});
+TORCH_API at::Tensor empty_strided(at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+TORCH_API at::Tensor empty_strided_symint(c10::SymIntArrayRef size, c10::SymIntArrayRef stride, at::TensorOptions options={});
+TORCH_API at::Tensor empty_strided_symint(c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/eq_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/eq_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3e2e51f3fac72594a88bcdb48712ecd256cfcc51
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/eq_meta_dispatch.h
@@ -0,0 +1,35 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 eq(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & eq_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & eq_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & eq_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor eq(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & eq_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & eq_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & eq_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/eq_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/eq_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..0088635bc480a1f36e29a3e68886846fe1e9d48a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/eq_ops.h
@@ -0,0 +1,89 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 eq__Scalar {
+  using schema = at::Tensor & (at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::eq_";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "eq_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Scalar & other);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Scalar & other);
+};
+
+struct TORCH_API eq__Tensor {
+  using schema = at::Tensor & (at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::eq_";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "eq_.Tensor(Tensor(a!) self, Tensor other) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Tensor & other);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Tensor & other);
+};
+
+struct TORCH_API eq_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 const char* name = "aten::eq";
+  static constexpr const char* overload_name = "Scalar_out";
+  static constexpr const char* schema_str = "eq.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+};
+
+struct TORCH_API eq_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 const char* name = "aten::eq";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "eq.Scalar(Tensor self, Scalar other) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Scalar & other);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & other);
+};
+
+struct TORCH_API eq_Tensor_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 const char* name = "aten::eq";
+  static constexpr const char* overload_name = "Tensor_out";
+  static constexpr const char* schema_str = "eq.Tensor_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 eq_Tensor {
+  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 const char* name = "aten::eq";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "eq.Tensor(Tensor self, Tensor other) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & other);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/equal.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/equal.h
new file mode 100644
index 0000000000000000000000000000000000000000..2c3d32036984ad03859514c2ea47ee260c9fb253
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/equal.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/equal_ops.h>
+
+namespace at {
+
+
+// aten::equal(Tensor self, Tensor other) -> bool
+inline bool equal(const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::equal::call(self, other);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/erf_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/erf_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d4b3cdd200e3d1d148764b4492d01a61c22a8d8d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/erf_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 erf(const at::Tensor & self);
+TORCH_API at::Tensor & erf_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & erf_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & erf_(at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/exp_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/exp_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ce09b1be18ae3e7d45413c0b99613a333a1b5549
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/exp_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 exp(const at::Tensor & self);
+TORCH_API at::Tensor & exp_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & exp_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & exp_(at::Tensor & self);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/exp_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/exp_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..af60ecafab90e9221ecc1bcf37d99840026c7cdb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/exp_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 exp(const at::Tensor & self);
+TORCH_API at::Tensor & exp_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & exp_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & exp_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/expand_as_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/expand_as_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a5a5c0234da2ba7727c7f148676b94cca7b45854
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/expand_as_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 expand_as(const at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/expm1_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/expm1_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..0964acdd234bfb461002aa5e2a762c87aea08e35
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/expm1_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 expm1(const at::Tensor & self);
+TORCH_API at::Tensor & expm1_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & expm1_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & expm1_(at::Tensor & self);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/eye_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/eye_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3db3f8489b564f006f59b3823fc8be73b778dbc5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/eye_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,35 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 eye(int64_t n, at::TensorOptions options={});
+TORCH_API at::Tensor eye(int64_t n, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+TORCH_API at::Tensor eye_symint(c10::SymInt n, at::TensorOptions options={});
+TORCH_API at::Tensor eye_symint(c10::SymInt n, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+TORCH_API at::Tensor eye(int64_t n, int64_t m, at::TensorOptions options={});
+TORCH_API at::Tensor eye(int64_t n, int64_t m, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+TORCH_API at::Tensor eye_symint(c10::SymInt n, c10::SymInt m, at::TensorOptions options={});
+TORCH_API at::Tensor eye_symint(c10::SymInt n, c10::SymInt m, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_channel_affine_cachemask.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_channel_affine_cachemask.h
new file mode 100644
index 0000000000000000000000000000000000000000..b290154ad2895f39cfddf97cd4e14b82604a2225
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_channel_affine_cachemask.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fake_quantize_per_channel_affine_cachemask_ops.h>
+
+namespace at {
+
+
+// aten::fake_quantize_per_channel_affine_cachemask(Tensor self, Tensor scale, Tensor zero_point, int axis, int quant_min, int quant_max) -> (Tensor output, Tensor mask)
+inline ::std::tuple<at::Tensor,at::Tensor> fake_quantize_per_channel_affine_cachemask(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis, int64_t quant_min, int64_t quant_max) {
+    return at::_ops::fake_quantize_per_channel_affine_cachemask::call(self, scale, zero_point, axis, quant_min, quant_max);
+}
+
+// aten::fake_quantize_per_channel_affine_cachemask.out(Tensor self, Tensor scale, Tensor zero_point, int axis, int quant_min, int quant_max, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))
+inline ::std::tuple<at::Tensor &,at::Tensor &> fake_quantize_per_channel_affine_cachemask_out(at::Tensor & out0, at::Tensor & out1, const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis, int64_t quant_min, int64_t quant_max) {
+    return at::_ops::fake_quantize_per_channel_affine_cachemask_out::call(self, scale, zero_point, axis, quant_min, quant_max, out0, out1);
+}
+// aten::fake_quantize_per_channel_affine_cachemask.out(Tensor self, Tensor scale, Tensor zero_point, int axis, int quant_min, int quant_max, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))
+inline ::std::tuple<at::Tensor &,at::Tensor &> fake_quantize_per_channel_affine_cachemask_outf(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis, int64_t quant_min, int64_t quant_max, at::Tensor & out0, at::Tensor & out1) {
+    return at::_ops::fake_quantize_per_channel_affine_cachemask_out::call(self, scale, zero_point, axis, quant_min, quant_max, out0, out1);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_channel_affine_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_channel_affine_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..eff92ec202a491cf02fc0b7e0eb07e8ff9953eff
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_channel_affine_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 fake_quantize_per_channel_affine(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis, int64_t quant_min, int64_t quant_max);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_channel_affine_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_channel_affine_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..031fe213560ba91a66249d7d21007d21ee9e568d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_channel_affine_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 fake_quantize_per_channel_affine {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, int64_t, int64_t, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fake_quantize_per_channel_affine";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "fake_quantize_per_channel_affine(Tensor self, Tensor scale, Tensor zero_point, int axis, int quant_min, int quant_max) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis, int64_t quant_min, int64_t quant_max);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t axis, int64_t quant_min, int64_t quant_max);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_cachemask_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_cachemask_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8d2f3ddfb40da53685803a233d3f4dedc119694a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_cachemask_backward_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 fake_quantize_per_tensor_affine_cachemask_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 const char* name = "aten::fake_quantize_per_tensor_affine_cachemask_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "fake_quantize_per_tensor_affine_cachemask_backward(Tensor grad, Tensor mask) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad, const at::Tensor & mask);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad, const at::Tensor & mask);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_cachemask_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_cachemask_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..96be82bca4b4f0bc49b3960279def1286443cf55
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_cachemask_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> fake_quantize_per_tensor_affine_cachemask_out(const at::Tensor & self, double scale, int64_t zero_point, int64_t quant_min, int64_t quant_max, at::Tensor & out0, at::Tensor & out1);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> fake_quantize_per_tensor_affine_cachemask(const at::Tensor & self, double scale, int64_t zero_point, int64_t quant_min, int64_t quant_max);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8979505e2c05a6d54f77cb6df814a9f7b3395b29
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 fake_quantize_per_tensor_affine(const at::Tensor & self, double scale, int64_t zero_point, int64_t quant_min, int64_t quant_max);
+TORCH_API at::Tensor fake_quantize_per_tensor_affine(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, int64_t quant_min, int64_t quant_max);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_fp16_weight.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_fp16_weight.h
new file mode 100644
index 0000000000000000000000000000000000000000..6e381fa3b9b06b6796e09186f55367aa9861755f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_fp16_weight.h
@@ -0,0 +1,41 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fbgemm_linear_fp16_weight_ops.h>
+
+namespace at {
+
+
+// aten::fbgemm_linear_fp16_weight(Tensor input, Tensor packed_weight, Tensor bias) -> Tensor
+inline at::Tensor fbgemm_linear_fp16_weight(const at::Tensor & input, const at::Tensor & packed_weight, const at::Tensor & bias) {
+    return at::_ops::fbgemm_linear_fp16_weight::call(input, packed_weight, bias);
+}
+
+// aten::fbgemm_linear_fp16_weight.out(Tensor input, Tensor packed_weight, Tensor bias, Tensor(a!) output) -> Tensor
+inline at::Tensor fbgemm_linear_fp16_weight(const at::Tensor & input, const at::Tensor & packed_weight, const at::Tensor & bias, at::Tensor & output) {
+    return at::_ops::fbgemm_linear_fp16_weight_out::call(input, packed_weight, bias, output);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_int8_weight_fp32_activation.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_int8_weight_fp32_activation.h
new file mode 100644
index 0000000000000000000000000000000000000000..3c47277ef771aef0358daa7b496992f26be68f83
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_int8_weight_fp32_activation.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fbgemm_linear_int8_weight_fp32_activation_ops.h>
+
+namespace at {
+
+
+// aten::fbgemm_linear_int8_weight_fp32_activation(Tensor input, Tensor weight, Tensor packed, Tensor col_offsets, Scalar weight_scale, Scalar weight_zero_point, Tensor bias) -> Tensor
+inline at::Tensor fbgemm_linear_int8_weight_fp32_activation(const at::Tensor & input, const at::Tensor & weight, const at::Tensor & packed, const at::Tensor & col_offsets, const at::Scalar & weight_scale, const at::Scalar & weight_zero_point, const at::Tensor & bias) {
+    return at::_ops::fbgemm_linear_int8_weight_fp32_activation::call(input, weight, packed, col_offsets, weight_scale, weight_zero_point, bias);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_int8_weight_fp32_activation_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_int8_weight_fp32_activation_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..e1c3cc6a0fd870e3980c284c651a59816a6aedd0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_int8_weight_fp32_activation_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API fbgemm_linear_int8_weight_fp32_activation {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Scalar &, const at::Scalar &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fbgemm_linear_int8_weight_fp32_activation";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "fbgemm_linear_int8_weight_fp32_activation(Tensor input, Tensor weight, Tensor packed, Tensor col_offsets, Scalar weight_scale, Scalar weight_zero_point, Tensor bias) -> Tensor";
+  static at::Tensor call(const at::Tensor & input, const at::Tensor & weight, const at::Tensor & packed, const at::Tensor & col_offsets, const at::Scalar & weight_scale, const at::Scalar & weight_zero_point, const at::Tensor & bias);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & weight, const at::Tensor & packed, const at::Tensor & col_offsets, const at::Scalar & weight_scale, const at::Scalar & weight_zero_point, const at::Tensor & bias);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_quantize_weight.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_quantize_weight.h
new file mode 100644
index 0000000000000000000000000000000000000000..a7c8bc07e257421934e29dadd95b5f763705d92b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_linear_quantize_weight.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fbgemm_linear_quantize_weight_ops.h>
+
+namespace at {
+
+
+// aten::fbgemm_linear_quantize_weight(Tensor input) -> (Tensor, Tensor, float, int)
+inline ::std::tuple<at::Tensor,at::Tensor,double,int64_t> fbgemm_linear_quantize_weight(const at::Tensor & input) {
+    return at::_ops::fbgemm_linear_quantize_weight::call(input);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_pack_gemm_matrix_fp16.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_pack_gemm_matrix_fp16.h
new file mode 100644
index 0000000000000000000000000000000000000000..bad86f420339b70d1e22ec7c959028a42859ab86
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_pack_gemm_matrix_fp16.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fbgemm_pack_gemm_matrix_fp16_ops.h>
+
+namespace at {
+
+
+// aten::fbgemm_pack_gemm_matrix_fp16(Tensor input) -> Tensor
+inline at::Tensor fbgemm_pack_gemm_matrix_fp16(const at::Tensor & input) {
+    return at::_ops::fbgemm_pack_gemm_matrix_fp16::call(input);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_pack_gemm_matrix_fp16_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_pack_gemm_matrix_fp16_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ff0a9fb1d3e79d6994759d01e7438959b9cb2ced
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_pack_gemm_matrix_fp16_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 fbgemm_pack_gemm_matrix_fp16(const at::Tensor & input);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_pack_quantized_matrix_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_pack_quantized_matrix_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3dabe1954216a8d61dd7be2c3ed0f8b692c32af6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fbgemm_pack_quantized_matrix_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 fbgemm_pack_quantized_matrix(const at::Tensor & input);
+TORCH_API at::Tensor fbgemm_pack_quantized_matrix(const at::Tensor & input, int64_t K, int64_t N);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/feature_alpha_dropout.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/feature_alpha_dropout.h
new file mode 100644
index 0000000000000000000000000000000000000000..58de921acd450a02f7e9039adf7386fb49de53c0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/feature_alpha_dropout.h
@@ -0,0 +1,41 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/feature_alpha_dropout_ops.h>
+
+namespace at {
+
+
+// aten::feature_alpha_dropout(Tensor input, float p, bool train) -> Tensor
+inline at::Tensor feature_alpha_dropout(const at::Tensor & input, double p, bool train) {
+    return at::_ops::feature_alpha_dropout::call(input, p, train);
+}
+
+// aten::feature_alpha_dropout_(Tensor(a!) self, float p, bool train) -> Tensor(a!)
+inline at::Tensor & feature_alpha_dropout_(at::Tensor & self, double p, bool train) {
+    return at::_ops::feature_alpha_dropout_::call(self, p, train);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/feature_dropout_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/feature_dropout_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..dac835f2370f22a5cd8ba693d45b5ceb5af5a2e1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/feature_dropout_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 feature_dropout(const at::Tensor & input, double p, bool train);
+TORCH_API at::Tensor & feature_dropout_(at::Tensor & self, double p, bool train);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fft2.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fft2.h
new file mode 100644
index 0000000000000000000000000000000000000000..2319ab578a6c2895fa024ffcca54b55ffbc224b6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fft2.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fft_fft2_ops.h>
+
+namespace at {
+
+
+// aten::fft_fft2(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None) -> Tensor
+inline at::Tensor fft_fft2(const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_fft2::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor fft_fft2(const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_fft2::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm);
+  }
+}
+
+// aten::fft_fft2(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None) -> Tensor
+inline at::Tensor fft_fft2_symint(const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_fft2::call(self, s, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor fft_fft2(const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_fft2::call(self, s, dim, norm);
+  }
+}
+
+// aten::fft_fft2.out(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_fft2_out(at::Tensor & out, const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_fft2_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & fft_fft2_out(at::Tensor & out, const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_fft2_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_fft2.out(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_fft2_outf(const at::Tensor & self, at::OptionalIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_fft2_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & fft_fft2_outf(const at::Tensor & self, at::OptionalIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_fft2_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_fft2.out(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_fft2_symint_out(at::Tensor & out, const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_fft2_out::call(self, s, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & fft_fft2_out(at::Tensor & out, const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_fft2_out::call(self, s, dim, norm, out);
+  }
+}
+
+// aten::fft_fft2.out(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_fft2_symint_outf(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_fft2_out::call(self, s, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & fft_fft2_outf(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_fft2_out::call(self, s, dim, norm, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fft2_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fft2_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e6cb5c6afeb0bd38fe972761530cc5fb0edc03ed
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fft2_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor fft_fft2(const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor fft_fft2_symint(const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_fft2_out(at::Tensor & out, const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_fft2_outf(const at::Tensor & self, at::OptionalIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+TORCH_API at::Tensor & fft_fft2_symint_out(at::Tensor & out, const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_fft2_symint_outf(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fft_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fft_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..1504e7e9802af3b71811343ce0e593d0bb049eb9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fft_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 fft_fft {
+  using schema = at::Tensor (const at::Tensor &, ::std::optional<c10::SymInt>, int64_t, ::std::optional<c10::string_view>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fft_fft";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "fft_fft(Tensor self, SymInt? n=None, int dim=-1, str? norm=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm);
+};
+
+struct TORCH_API fft_fft_out {
+  using schema = at::Tensor & (const at::Tensor &, ::std::optional<c10::SymInt>, int64_t, ::std::optional<c10::string_view>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fft_fft";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "fft_fft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fftshift_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fftshift_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..ff6019ee718a61dbc339696f9eb50548a853740a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_fftshift_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_fftshift(const at::Tensor & self, at::OptionalIntArrayRef dim=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_hfftn_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_hfftn_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..3d906c1c182582e983f4d782966f6ab353e28ad7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_hfftn_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_hfftn_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_hfftn_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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifft.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifft.h
new file mode 100644
index 0000000000000000000000000000000000000000..d81a8e8773c10a283a7285dcbadeff4731acee4e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifft.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fft_ifft_ops.h>
+
+namespace at {
+
+
+// aten::fft_ifft(Tensor self, SymInt? n=None, int dim=-1, str? norm=None) -> Tensor
+inline at::Tensor fft_ifft(const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ifft::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor fft_ifft(const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ifft::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm);
+  }
+}
+
+// aten::fft_ifft(Tensor self, SymInt? n=None, int dim=-1, str? norm=None) -> Tensor
+inline at::Tensor fft_ifft_symint(const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ifft::call(self, n, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor fft_ifft(const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ifft::call(self, n, dim, norm);
+  }
+}
+
+// aten::fft_ifft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ifft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ifft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & fft_ifft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ifft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_ifft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ifft_outf(const at::Tensor & self, ::std::optional<int64_t> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ifft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & fft_ifft_outf(const at::Tensor & self, ::std::optional<int64_t> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ifft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_ifft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ifft_symint_out(at::Tensor & out, const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ifft_out::call(self, n, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & fft_ifft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ifft_out::call(self, n, dim, norm, out);
+  }
+}
+
+// aten::fft_ifft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ifft_symint_outf(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ifft_out::call(self, n, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & fft_ifft_outf(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ifft_out::call(self, n, dim, norm, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifft2_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifft2_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..3e0f18274f94e7c5d2c8830626f51abb7c8e6b1d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifft2_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 fft_ifft2 {
+  using schema = at::Tensor (const at::Tensor &, at::OptionalSymIntArrayRef, at::IntArrayRef, ::std::optional<c10::string_view>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fft_ifft2";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "fft_ifft2(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm);
+};
+
+struct TORCH_API fft_ifft2_out {
+  using schema = at::Tensor & (const at::Tensor &, at::OptionalSymIntArrayRef, at::IntArrayRef, ::std::optional<c10::string_view>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fft_ifft2";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "fft_ifft2.out(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifftn_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifftn_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b99cde3f4b1e2954cb6203cea74eeeb278d3c100
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifftn_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifftshift_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifftshift_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8b01459753d7e46527d9b4f1766dbba312de2ccb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ifftshift_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 fft_ifftshift {
+  using schema = at::Tensor (const at::Tensor &, at::OptionalIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fft_ifftshift";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "fft_ifftshift(Tensor self, int[1]? dim=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::OptionalIntArrayRef dim);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::OptionalIntArrayRef dim);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft.h
new file mode 100644
index 0000000000000000000000000000000000000000..345381b77a6610c26760398ff51de5ceb4d0fc9d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fft_ihfft_ops.h>
+
+namespace at {
+
+
+// aten::fft_ihfft(Tensor self, SymInt? n=None, int dim=-1, str? norm=None) -> Tensor
+inline at::Tensor fft_ihfft(const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor fft_ihfft(const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm);
+  }
+}
+
+// aten::fft_ihfft(Tensor self, SymInt? n=None, int dim=-1, str? norm=None) -> Tensor
+inline at::Tensor fft_ihfft_symint(const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft::call(self, n, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor fft_ihfft(const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft::call(self, n, dim, norm);
+  }
+}
+
+// aten::fft_ihfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & fft_ihfft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_ihfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft_outf(const at::Tensor & self, ::std::optional<int64_t> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & fft_ihfft_outf(const at::Tensor & self, ::std::optional<int64_t> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_ihfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft_symint_out(at::Tensor & out, const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft_out::call(self, n, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & fft_ihfft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft_out::call(self, n, dim, norm, out);
+  }
+}
+
+// aten::fft_ihfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft_symint_outf(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft_out::call(self, n, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & fft_ihfft_outf(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft_out::call(self, n, dim, norm, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft2.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft2.h
new file mode 100644
index 0000000000000000000000000000000000000000..ac32165aceba9cf8fd387fc860c967250ecb5e08
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft2.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fft_ihfft2_ops.h>
+
+namespace at {
+
+
+// aten::fft_ihfft2(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None) -> Tensor
+inline at::Tensor fft_ihfft2(const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft2::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor fft_ihfft2(const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft2::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm);
+  }
+}
+
+// aten::fft_ihfft2(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None) -> Tensor
+inline at::Tensor fft_ihfft2_symint(const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft2::call(self, s, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor fft_ihfft2(const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft2::call(self, s, dim, norm);
+  }
+}
+
+// aten::fft_ihfft2.out(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft2_out(at::Tensor & out, const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft2_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & fft_ihfft2_out(at::Tensor & out, const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft2_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_ihfft2.out(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft2_outf(const at::Tensor & self, at::OptionalIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft2_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & fft_ihfft2_outf(const at::Tensor & self, at::OptionalIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft2_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_ihfft2.out(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft2_symint_out(at::Tensor & out, const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft2_out::call(self, s, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & fft_ihfft2_out(at::Tensor & out, const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_ihfft2_out::call(self, s, dim, norm, out);
+  }
+}
+
+// aten::fft_ihfft2.out(Tensor self, SymInt[1]? s=None, int[1] dim=[-2,-1], str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_ihfft2_symint_outf(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft2_out::call(self, s, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & fft_ihfft2_outf(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_ihfft2_out::call(self, s, dim, norm, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft2_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft2_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..a3caae440ca4e78ae63b3dc7e80c2cf8c442ce75
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft2_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_ihfft2_symint(const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_ihfft2_symint_out(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..36f1361d5a8fbe3703f0d7d653bb9506e26e0242
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_ihfft_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor fft_ihfft(const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor fft_ihfft_symint(const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_ihfft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_ihfft_outf(const at::Tensor & self, ::std::optional<int64_t> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+TORCH_API at::Tensor & fft_ihfft_symint_out(at::Tensor & out, const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_ihfft_symint_outf(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_irfft2_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_irfft2_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..627690f4d12706828a3c1afca7557cc576ae9d2c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_irfft2_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_irfft2_symint(const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_irfft2_symint_out(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_irfftn.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_irfftn.h
new file mode 100644
index 0000000000000000000000000000000000000000..7a41423359b257505e01a28f62803db02ec835f2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_irfftn.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fft_irfftn_ops.h>
+
+namespace at {
+
+
+// aten::fft_irfftn(Tensor self, SymInt[1]? s=None, int[1]? dim=None, str? norm=None) -> Tensor
+inline at::Tensor fft_irfftn(const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::OptionalIntArrayRef dim=::std::nullopt, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_irfftn::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor fft_irfftn(const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::OptionalIntArrayRef dim=::std::nullopt, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_irfftn::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm);
+  }
+}
+
+// aten::fft_irfftn(Tensor self, SymInt[1]? s=None, int[1]? dim=None, str? norm=None) -> Tensor
+inline at::Tensor fft_irfftn_symint(const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::OptionalIntArrayRef dim=::std::nullopt, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_irfftn::call(self, s, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor fft_irfftn(const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::OptionalIntArrayRef dim=::std::nullopt, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_irfftn::call(self, s, dim, norm);
+  }
+}
+
+// aten::fft_irfftn.out(Tensor self, SymInt[1]? s=None, int[1]? dim=None, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_irfftn_out(at::Tensor & out, const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::OptionalIntArrayRef dim=::std::nullopt, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_irfftn_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & fft_irfftn_out(at::Tensor & out, const at::Tensor & self, at::OptionalIntArrayRef s=::std::nullopt, at::OptionalIntArrayRef dim=::std::nullopt, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_irfftn_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_irfftn.out(Tensor self, SymInt[1]? s=None, int[1]? dim=None, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_irfftn_outf(const at::Tensor & self, at::OptionalIntArrayRef s, at::OptionalIntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_irfftn_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & fft_irfftn_outf(const at::Tensor & self, at::OptionalIntArrayRef s, at::OptionalIntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_irfftn_out::call(self, s.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*s)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_irfftn.out(Tensor self, SymInt[1]? s=None, int[1]? dim=None, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_irfftn_symint_out(at::Tensor & out, const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::OptionalIntArrayRef dim=::std::nullopt, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_irfftn_out::call(self, s, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & fft_irfftn_out(at::Tensor & out, const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::OptionalIntArrayRef dim=::std::nullopt, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_irfftn_out::call(self, s, dim, norm, out);
+  }
+}
+
+// aten::fft_irfftn.out(Tensor self, SymInt[1]? s=None, int[1]? dim=None, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_irfftn_symint_outf(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::OptionalIntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_irfftn_out::call(self, s, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & fft_irfftn_outf(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::OptionalIntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_irfftn_out::call(self, s, dim, norm, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_irfftn_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_irfftn_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..e6d7703a1bdb820f92ba007c139dee86b1e47674
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_irfftn_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 fft_irfftn {
+  using schema = at::Tensor (const at::Tensor &, at::OptionalSymIntArrayRef, at::OptionalIntArrayRef, ::std::optional<c10::string_view>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fft_irfftn";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "fft_irfftn(Tensor self, SymInt[1]? s=None, int[1]? dim=None, str? norm=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::OptionalIntArrayRef dim, ::std::optional<c10::string_view> norm);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::OptionalSymIntArrayRef s, at::OptionalIntArrayRef dim, ::std::optional<c10::string_view> norm);
+};
+
+struct TORCH_API fft_irfftn_out {
+  using schema = at::Tensor & (const at::Tensor &, at::OptionalSymIntArrayRef, at::OptionalIntArrayRef, ::std::optional<c10::string_view>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fft_irfftn";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "fft_irfftn.out(Tensor self, SymInt[1]? s=None, int[1]? dim=None, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::OptionalIntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::OptionalSymIntArrayRef s, at::OptionalIntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_rfft2_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_rfft2_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..5831542e7b560a85b7e748dfb9068089424aa2e4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fft_rfft2_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_rfft2_symint(const at::Tensor & self, at::OptionalSymIntArrayRef s=::std::nullopt, at::IntArrayRef dim={-2,-1}, ::std::optional<c10::string_view> norm=::std::nullopt);
+TORCH_API at::Tensor & fft_rfft2_symint_out(const at::Tensor & self, at::OptionalSymIntArrayRef s, at::IntArrayRef dim, ::std::optional<c10::string_view> norm, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fill_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fill_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..451b557b15da723f56a0a2a7058ac703fe7a7028
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fill_cuda_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_(at::Tensor & self, const at::Scalar & value);
+TORCH_API at::Tensor & fill_(at::Tensor & self, const at::Tensor & value);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/flip_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/flip_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ec060a9b54b9c39955f038b5dfb9529b441e7c66
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/flip_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & flip_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef dims);
+TORCH_API at::Tensor & flip_outf(const at::Tensor & self, at::IntArrayRef dims, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fliplr_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fliplr_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..9996687ec8659611de6293a87efed24bace6d0ac
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fliplr_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 fliplr {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fliplr";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "fliplr(Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..103a73372bdf715995b8f17b852e6008d2db3b80
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 floor(const at::Tensor & self);
+TORCH_API at::Tensor & floor_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & floor_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & floor_(at::Tensor & self);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_divide_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_divide_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..fdfade79e613308e978c7a33e9925d16d4be2e66
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_divide_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 floor_divide(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & floor_divide_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & floor_divide_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & floor_divide_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_divide_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_divide_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..30385dc4427bd11e3fb1e79fd9ff4e92b81a4414
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_divide_native.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 floor_divide(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & floor_divide_out(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & floor_divide_(at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor floor_divide_sparse(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & floor_divide_out_sparse_zerodim(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & floor_divide_sparse_(at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor floor_divide(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & floor_divide_Scalar_out(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & floor_divide_(at::Tensor & self, const at::Scalar & other);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..512b6cd0b8aa4a4a763e28bbd837530e81d5bc36
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/floor_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 floor(const at::Tensor & self);
+TORCH_API at::Tensor & floor_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & floor_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & floor_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmin.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmin.h
new file mode 100644
index 0000000000000000000000000000000000000000..3e68f0f179dede53b9892f6d1d3267dae056f8fd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmin.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fmin_ops.h>
+
+namespace at {
+
+
+// aten::fmin(Tensor self, Tensor other) -> Tensor
+inline at::Tensor fmin(const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::fmin::call(self, other);
+}
+
+// aten::fmin.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fmin_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::fmin_out::call(self, other, out);
+}
+// aten::fmin.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fmin_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out) {
+    return at::_ops::fmin_out::call(self, other, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmin_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmin_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..193b63bca96836df420e1cdebac54bb0dec2e4f8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmin_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/fmin_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_fmin_out : public at::meta::structured_fmin {
+void impl(const at::Tensor & self, const at::Tensor & other, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmod.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmod.h
new file mode 100644
index 0000000000000000000000000000000000000000..cee471d2a957037f8f66d7d240d0f62ae8cd2ff1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmod.h
@@ -0,0 +1,59 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fmod_ops.h>
+
+namespace at {
+
+
+// aten::fmod.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fmod_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other) {
+    return at::_ops::fmod_Scalar_out::call(self, other, out);
+}
+// aten::fmod.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fmod_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out) {
+    return at::_ops::fmod_Scalar_out::call(self, other, out);
+}
+
+// aten::fmod.Scalar(Tensor self, Scalar other) -> Tensor
+inline at::Tensor fmod(const at::Tensor & self, const at::Scalar & other) {
+    return at::_ops::fmod_Scalar::call(self, other);
+}
+
+// aten::fmod.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fmod_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::fmod_Tensor_out::call(self, other, out);
+}
+// aten::fmod.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fmod_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out) {
+    return at::_ops::fmod_Tensor_out::call(self, other, out);
+}
+
+// aten::fmod.Tensor(Tensor self, Tensor other) -> Tensor
+inline at::Tensor fmod(const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::fmod_Tensor::call(self, other);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmod_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmod_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c7836eb9d1931b1b3161ba5a9ec589171d61a525
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmod_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 fmod(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & fmod_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & fmod_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & fmod_(at::Tensor & self, const at::Scalar & other);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmod_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmod_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c2d0fd152c8a2bd5f2771fcd5b569246573b59e2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fmod_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API at::Tensor fmod(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & fmod_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & fmod_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & fmod_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/frac_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/frac_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..be13b5595252c5fea625c6d64ed46c2d12749e2e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/frac_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 frac(const at::Tensor & self);
+TORCH_API at::Tensor & frac_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & frac_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & frac_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool2d.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool2d.h
new file mode 100644
index 0000000000000000000000000000000000000000..4838184068cd9e2d14f7e73d8497759c741c55cf
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool2d.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/fractional_max_pool2d_ops.h>
+
+namespace at {
+
+
+// aten::fractional_max_pool2d.output(Tensor self, int[2] kernel_size, int[2] output_size, Tensor random_samples, *, Tensor(a!) output, Tensor(b!) indices) -> (Tensor(a!), Tensor(b!))
+inline ::std::tuple<at::Tensor &,at::Tensor &> fractional_max_pool2d_out(at::Tensor & output, at::Tensor & indices, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples) {
+    return at::_ops::fractional_max_pool2d_output::call(self, kernel_size, output_size, random_samples, output, indices);
+}
+// aten::fractional_max_pool2d.output(Tensor self, int[2] kernel_size, int[2] output_size, Tensor random_samples, *, Tensor(a!) output, Tensor(b!) indices) -> (Tensor(a!), Tensor(b!))
+inline ::std::tuple<at::Tensor &,at::Tensor &> fractional_max_pool2d_outf(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples, at::Tensor & output, at::Tensor & indices) {
+    return at::_ops::fractional_max_pool2d_output::call(self, kernel_size, output_size, random_samples, output, indices);
+}
+
+// aten::fractional_max_pool2d(Tensor self, int[2] kernel_size, int[2] output_size, Tensor random_samples) -> (Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor> fractional_max_pool2d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples) {
+    return at::_ops::fractional_max_pool2d::call(self, kernel_size, output_size, random_samples);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool2d_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool2d_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..e404de2fcce0eb980bf9c4a4b0fcf937236e8e26
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool2d_backward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API fractional_max_pool2d_backward_grad_input {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, at::IntArrayRef, at::IntArrayRef, const at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fractional_max_pool2d_backward";
+  static constexpr const char* overload_name = "grad_input";
+  static constexpr const char* schema_str = "fractional_max_pool2d_backward.grad_input(Tensor grad_output, Tensor self, int[2] kernel_size, int[2] output_size, Tensor indices, *, Tensor(a!) grad_input) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & indices, at::Tensor & grad_input);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & indices, at::Tensor & grad_input);
+};
+
+struct TORCH_API fractional_max_pool2d_backward {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, at::IntArrayRef, at::IntArrayRef, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fractional_max_pool2d_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "fractional_max_pool2d_backward(Tensor grad_output, Tensor self, int[2] kernel_size, int[2] output_size, Tensor indices) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & indices);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & indices);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool2d_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool2d_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..43a6510dfadbea3159b72487e1768b0a2d002d2c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool2d_native.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/fractional_max_pool2d_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_fractional_max_pool2d_out_cpu : public at::meta::structured_fractional_max_pool2d {
+void impl(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples, const at::Tensor & output, const at::Tensor & indices);
+};
+struct TORCH_API structured_fractional_max_pool2d_out_cuda : public at::meta::structured_fractional_max_pool2d {
+void impl(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples, const at::Tensor & output, const at::Tensor & indices);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..68a287fa6cfd7739bd1e293fff1b7a76711438d2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_backward_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 fractional_max_pool3d_backward_cpu(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & indices);
+TORCH_API at::Tensor & fractional_max_pool3d_backward_out_cpu(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & indices, at::Tensor & grad_input);
+TORCH_API at::Tensor fractional_max_pool3d_backward_cuda(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & indices);
+TORCH_API at::Tensor & fractional_max_pool3d_backward_out_cuda(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & indices, at::Tensor & grad_input);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3f45ee686b25c6e69f6f4ce58e9c65476a985c7e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,at::Tensor> fractional_max_pool3d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..1ea60ce5446bc5fbb63a4ca89af5be5d23c5e1ae
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_native.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/fractional_max_pool3d_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_fractional_max_pool3d_out_cpu : public at::meta::structured_fractional_max_pool3d {
+void impl(const at::Tensor & self, int64_t poolSizeT, int64_t poolSizeH, int64_t poolSizeW, int64_t outputT, int64_t outputH, int64_t outputW, const at::Tensor & random_samples, int64_t numBatch, int64_t numPlanes, int64_t inputT, int64_t inputH, int64_t inputW, const at::Tensor & output, const at::Tensor & indices);
+};
+struct TORCH_API structured_fractional_max_pool3d_out_cuda : public at::meta::structured_fractional_max_pool3d {
+void impl(const at::Tensor & self, int64_t poolSizeT, int64_t poolSizeH, int64_t poolSizeW, int64_t outputT, int64_t outputH, int64_t outputW, const at::Tensor & random_samples, int64_t numBatch, int64_t numPlanes, int64_t inputT, int64_t inputH, int64_t inputW, const at::Tensor & output, const at::Tensor & indices);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..220b72dede09b2520a3034d1daef2f50bea7437a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/fractional_max_pool3d_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API fractional_max_pool3d_output {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, at::IntArrayRef, at::IntArrayRef, const at::Tensor &, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fractional_max_pool3d";
+  static constexpr const char* overload_name = "output";
+  static constexpr const char* schema_str = "fractional_max_pool3d.output(Tensor self, int[3] kernel_size, int[3] output_size, Tensor random_samples, *, Tensor(a!) output, Tensor(b!) indices) -> (Tensor(a!), Tensor(b!))";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples, at::Tensor & output, at::Tensor & indices);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples, at::Tensor & output, at::Tensor & indices);
+};
+
+struct TORCH_API fractional_max_pool3d {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, at::IntArrayRef, at::IntArrayRef, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::fractional_max_pool3d";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "fractional_max_pool3d(Tensor self, int[3] kernel_size, int[3] output_size, Tensor random_samples) -> (Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/full_like.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/full_like.h
new file mode 100644
index 0000000000000000000000000000000000000000..388105a0093d63455ce8c8dcee6294aa95e37db4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/full_like.h
@@ -0,0 +1,49 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/full_like_ops.h>
+
+namespace at {
+
+
+// aten::full_like(Tensor self, Scalar fill_value, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor
+inline at::Tensor full_like(const at::Tensor & self, const at::Scalar & fill_value, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt) {
+    return at::_ops::full_like::call(self, fill_value, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt(), c10::impl::check_tensor_options_and_extract_memory_format(options, memory_format));
+}
+// aten::full_like(Tensor self, Scalar fill_value, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor
+inline at::Tensor full_like(const at::Tensor & self, const at::Scalar & fill_value, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format) {
+    return at::_ops::full_like::call(self, fill_value, dtype, layout, device, pin_memory, memory_format);
+}
+
+// aten::full_like.out(Tensor self, Scalar fill_value, *, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & full_like_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & fill_value, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt) {
+    return at::_ops::full_like_out::call(self, fill_value, memory_format, out);
+}
+// aten::full_like.out(Tensor self, Scalar fill_value, *, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & full_like_outf(const at::Tensor & self, const at::Scalar & fill_value, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out) {
+    return at::_ops::full_like_out::call(self, fill_value, memory_format, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/full_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/full_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..3e8ff00a112fb9bf29fb4cb357ca9dd4936f2df3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/full_ops.h
@@ -0,0 +1,67 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 full_names {
+  using schema = at::Tensor (at::IntArrayRef, const at::Scalar &, ::std::optional<at::DimnameList>, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::full";
+  static constexpr const char* overload_name = "names";
+  static constexpr const char* schema_str = "full.names(int[] size, Scalar fill_value, *, Dimname[]? names, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor";
+  static at::Tensor call(at::IntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::DimnameList> names, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, at::IntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::DimnameList> names, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+struct TORCH_API full {
+  using schema = at::Tensor (c10::SymIntArrayRef, const at::Scalar &, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::full";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "full(SymInt[] size, Scalar fill_value, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor";
+  static at::Tensor call(c10::SymIntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, c10::SymIntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+struct TORCH_API full_out {
+  using schema = at::Tensor & (c10::SymIntArrayRef, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::full";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "full.out(SymInt[] size, Scalar fill_value, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(c10::SymIntArrayRef size, const at::Scalar & fill_value, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, c10::SymIntArrayRef size, const at::Scalar & fill_value, at::Tensor & out);
+};
+
+struct TORCH_API full_names_out {
+  using schema = at::Tensor & (at::IntArrayRef, const at::Scalar &, ::std::optional<at::DimnameList>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::full";
+  static constexpr const char* overload_name = "names_out";
+  static constexpr const char* schema_str = "full.names_out(int[] size, Scalar fill_value, *, Dimname[]? names, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(at::IntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::DimnameList> names, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::IntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::DimnameList> names, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..a1ee6bc211ff66ac34602b22d540b7e5c76300fe
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_backward_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 gather_backward(const at::Tensor & grad, const at::Tensor & self, int64_t dim, const at::Tensor & index, bool sparse_grad);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..03c272f37d288a1ee66e366db902e749c9b18878
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_backward_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 gather_backward {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, int64_t, const at::Tensor &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::gather_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "gather_backward(Tensor grad, Tensor self, int dim, Tensor index, bool sparse_grad) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad, const at::Tensor & self, int64_t dim, const at::Tensor & index, bool sparse_grad);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad, const at::Tensor & self, int64_t dim, const at::Tensor & index, bool sparse_grad);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..2938c5e62e0737ec7e0c9dcf46c57c8ad1af73ac
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 gather(const at::Tensor & self, int64_t dim, const at::Tensor & index, bool sparse_grad=false);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e86bf4fc214cd0640a507695695ce5d0ea55a087
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gather_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 gather(const at::Tensor & self, int64_t dim, const at::Tensor & index, bool sparse_grad=false);
+TORCH_API at::Tensor & gather_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index, bool sparse_grad=false);
+TORCH_API at::Tensor & gather_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, bool sparse_grad, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ge_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ge_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d55c748171314102282ebfea25c3e721749a6f1c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ge_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ge(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & ge_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor ge(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & ge_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu.h
new file mode 100644
index 0000000000000000000000000000000000000000..5d856781359fad53703267957320a680647515ea
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/gelu_ops.h>
+
+namespace at {
+
+
+// aten::gelu.out(Tensor self, *, str approximate='none', Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & gelu_out(at::Tensor & out, const at::Tensor & self, c10::string_view approximate="none") {
+    return at::_ops::gelu_out::call(self, approximate, out);
+}
+// aten::gelu.out(Tensor self, *, str approximate='none', Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & gelu_outf(const at::Tensor & self, c10::string_view approximate, at::Tensor & out) {
+    return at::_ops::gelu_out::call(self, approximate, out);
+}
+
+// aten::gelu_(Tensor(a!) self, *, str approximate='none') -> Tensor(a!)
+inline at::Tensor & gelu_(at::Tensor & self, c10::string_view approximate="none") {
+    return at::_ops::gelu_::call(self, approximate);
+}
+
+// aten::gelu(Tensor self, *, str approximate='none') -> Tensor
+inline at::Tensor gelu(const at::Tensor & self, c10::string_view approximate="none") {
+    return at::_ops::gelu::call(self, approximate);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_backward.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..d05a76c3b5c04dd8fc85eabf11f6108441738663
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_backward.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/gelu_backward_ops.h>
+
+namespace at {
+
+
+// aten::gelu_backward.grad_input(Tensor grad_output, Tensor self, *, str approximate='none', Tensor(a!) grad_input) -> Tensor(a!)
+inline at::Tensor & gelu_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, c10::string_view approximate="none") {
+    return at::_ops::gelu_backward_grad_input::call(grad_output, self, approximate, grad_input);
+}
+// aten::gelu_backward.grad_input(Tensor grad_output, Tensor self, *, str approximate='none', Tensor(a!) grad_input) -> Tensor(a!)
+inline at::Tensor & gelu_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, c10::string_view approximate, at::Tensor & grad_input) {
+    return at::_ops::gelu_backward_grad_input::call(grad_output, self, approximate, grad_input);
+}
+
+// aten::gelu_backward(Tensor grad_output, Tensor self, *, str approximate='none') -> Tensor
+inline at::Tensor gelu_backward(const at::Tensor & grad_output, const at::Tensor & self, c10::string_view approximate="none") {
+    return at::_ops::gelu_backward::call(grad_output, self, approximate);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f6139ff2355fc9f04afb4810ee57de5ab542cae1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4698e6a5ea07cd8ffbd26afdb8edc55e254fe97a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 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 meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..f1efd60bbce118fb1b0bbee075366e2a30cb0f50
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gelu_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 gelu_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::string_view, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::gelu";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "gelu.out(Tensor self, *, str approximate='none', Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::string_view approximate, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::string_view approximate, at::Tensor & out);
+};
+
+struct TORCH_API gelu_ {
+  using schema = at::Tensor & (at::Tensor &, c10::string_view);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::gelu_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "gelu_(Tensor(a!) self, *, str approximate='none') -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, c10::string_view approximate);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, c10::string_view approximate);
+};
+
+struct TORCH_API gelu {
+  using schema = at::Tensor (const at::Tensor &, c10::string_view);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::gelu";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "gelu(Tensor self, *, str approximate='none') -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::string_view approximate);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::string_view approximate);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8479f9f11a18bf01f994889c382dff35b1af7e9e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 glu_backward(const at::Tensor & grad_output, const at::Tensor & self, int64_t dim);
+TORCH_API at::Tensor & glu_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, int64_t dim);
+TORCH_API at::Tensor & glu_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, int64_t dim, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp.h
new file mode 100644
index 0000000000000000000000000000000000000000..70862c6182ef346df544f28c5446df64105bc1ad
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/glu_backward_jvp_ops.h>
+
+namespace at {
+
+
+// aten::glu_backward_jvp(Tensor grad_x, Tensor grad_glu, Tensor x, Tensor dgrad_glu, Tensor dx, int dim) -> Tensor
+inline at::Tensor glu_backward_jvp(const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim) {
+    return at::_ops::glu_backward_jvp::call(grad_x, grad_glu, x, dgrad_glu, dx, dim);
+}
+
+// aten::glu_backward_jvp.out(Tensor grad_x, Tensor grad_glu, Tensor x, Tensor dgrad_glu, Tensor dx, int dim, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & glu_backward_jvp_out(at::Tensor & out, const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim) {
+    return at::_ops::glu_backward_jvp_out::call(grad_x, grad_glu, x, dgrad_glu, dx, dim, out);
+}
+// aten::glu_backward_jvp.out(Tensor grad_x, Tensor grad_glu, Tensor x, Tensor dgrad_glu, Tensor dx, int dim, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & glu_backward_jvp_outf(const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim, at::Tensor & out) {
+    return at::_ops::glu_backward_jvp_out::call(grad_x, grad_glu, x, dgrad_glu, dx, dim, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d1798fe8eeba666c48265b639dee858e34ce3804
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor glu_backward_jvp(const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..1dbcc797b5476112c904ec49b9ee0b26214cf32c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & glu_backward_jvp_out(const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim, at::Tensor & out);
+TORCH_API at::Tensor glu_backward_jvp(const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..dcac0157b6f0e5b9d9021f63458cdbc37248ba05
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_backward_jvp_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 glu_backward_jvp {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::glu_backward_jvp";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "glu_backward_jvp(Tensor grad_x, Tensor grad_glu, Tensor x, Tensor dgrad_glu, Tensor dx, int dim) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim);
+};
+
+struct TORCH_API glu_backward_jvp_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, int64_t, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::glu_backward_jvp";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "glu_backward_jvp.out(Tensor grad_x, Tensor grad_glu, Tensor x, Tensor dgrad_glu, Tensor dx, int dim, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4e4a7fcd0447011ba8af68c5521cf529793e8f07
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor glu(const at::Tensor & self, int64_t dim=-1);
+TORCH_API at::Tensor & glu_out(at::Tensor & out, const at::Tensor & self, int64_t dim=-1);
+TORCH_API at::Tensor & glu_outf(const at::Tensor & self, int64_t dim, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_jvp.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_jvp.h
new file mode 100644
index 0000000000000000000000000000000000000000..e4066806c5bb0003d2bf01a2271bee0eb4913915
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_jvp.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/glu_jvp_ops.h>
+
+namespace at {
+
+
+// aten::glu_jvp(Tensor glu, Tensor x, Tensor dx, int dim) -> Tensor
+inline at::Tensor glu_jvp(const at::Tensor & glu, const at::Tensor & x, const at::Tensor & dx, int64_t dim) {
+    return at::_ops::glu_jvp::call(glu, x, dx, dim);
+}
+
+// aten::glu_jvp.out(Tensor glu, Tensor x, Tensor dx, int dim, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & glu_jvp_out(at::Tensor & out, const at::Tensor & glu, const at::Tensor & x, const at::Tensor & dx, int64_t dim) {
+    return at::_ops::glu_jvp_out::call(glu, x, dx, dim, out);
+}
+// aten::glu_jvp.out(Tensor glu, Tensor x, Tensor dx, int dim, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & glu_jvp_outf(const at::Tensor & glu, const at::Tensor & x, const at::Tensor & dx, int64_t dim, at::Tensor & out) {
+    return at::_ops::glu_jvp_out::call(glu, x, dx, dim, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_jvp_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_jvp_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..62b26abe272cabf1f49a56023c8e388eee6545d7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/glu_jvp_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 glu_jvp(const at::Tensor & glu, const at::Tensor & x, const at::Tensor & dx, int64_t dim);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gradient_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gradient_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..2b09ae9f0aa196c1a39f2124d6991e1eb5616ccf
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/gradient_native.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::vector<at::Tensor> gradient(const at::Tensor & self, const ::std::optional<at::Scalar> & spacing=::std::nullopt, ::std::optional<int64_t> dim=::std::nullopt, int64_t edge_order=1);
+TORCH_API ::std::vector<at::Tensor> gradient(const at::Tensor & self, const at::Scalar & spacing, at::IntArrayRef dim, int64_t edge_order=1);
+TORCH_API ::std::vector<at::Tensor> gradient(const at::Tensor & self, at::IntArrayRef dim, int64_t edge_order=1);
+TORCH_API ::std::vector<at::Tensor> gradient(const at::Tensor & self, at::ArrayRef<at::Scalar> spacing, ::std::optional<int64_t> dim=::std::nullopt, int64_t edge_order=1);
+TORCH_API ::std::vector<at::Tensor> gradient(const at::Tensor & self, at::ArrayRef<at::Scalar> spacing, at::IntArrayRef dim, int64_t edge_order=1);
+TORCH_API ::std::vector<at::Tensor> gradient(const at::Tensor & self, at::TensorList spacing, ::std::optional<int64_t> dim=::std::nullopt, int64_t edge_order=1);
+TORCH_API ::std::vector<at::Tensor> gradient(const at::Tensor & self, at::TensorList spacing, at::IntArrayRef dim, int64_t edge_order=1);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/greater_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/greater_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4c356914ee399e7cc1677bb2931d8c4d04c0cf61
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/greater_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,35 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 greater(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & greater_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & greater_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & greater_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor greater(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & greater_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & greater_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & greater_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/greater_equal_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/greater_equal_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..73916445901e41e4f6069a249d81e4e3c19d25b7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/greater_equal_ops.h
@@ -0,0 +1,89 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 greater_equal_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 const char* name = "aten::greater_equal";
+  static constexpr const char* overload_name = "Scalar_out";
+  static constexpr const char* schema_str = "greater_equal.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+};
+
+struct TORCH_API greater_equal_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 const char* name = "aten::greater_equal";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "greater_equal.Scalar(Tensor self, Scalar other) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Scalar & other);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & other);
+};
+
+struct TORCH_API greater_equal_Tensor_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 const char* name = "aten::greater_equal";
+  static constexpr const char* overload_name = "Tensor_out";
+  static constexpr const char* schema_str = "greater_equal.Tensor_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 greater_equal_Tensor {
+  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 const char* name = "aten::greater_equal";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "greater_equal.Tensor(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 greater_equal__Scalar {
+  using schema = at::Tensor & (at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::greater_equal_";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "greater_equal_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Scalar & other);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Scalar & other);
+};
+
+struct TORCH_API greater_equal__Tensor {
+  using schema = at::Tensor & (at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::greater_equal_";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "greater_equal_.Tensor(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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_2d_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_2d_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..048ef9bde9d6e32f38942ad282d2b5e557524986
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_2d_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_3d_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_3d_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f3de7e3ae1cdfab5b1af6ff5ceceb438e8e00216
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_3d_backward_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> grid_sampler_3d_backward(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners, ::std::array<bool,2> output_mask);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_3d_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_3d_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..752b214c8dfe3137d0b2bda4b3dd5f71dac99e41
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_3d_backward_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> grid_sampler_3d_backward_out(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners, ::std::array<bool,2> output_mask, at::Tensor & out0, at::Tensor & out1);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> grid_sampler_3d_backward_cpu(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners, ::std::array<bool,2> output_mask);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> grid_sampler_3d_backward_cuda(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners, ::std::array<bool,2> output_mask);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_3d_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_3d_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..032354a154730f540a81d06c2cfb961e8acdc776
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_3d_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 grid_sampler_3d(const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b6432269a09d1def017226ac14fd0739c05aa144
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 grid_sampler(const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..2e3ba04a6c92de9c9b84de23b137509c27106ddd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 grid_sampler(const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..31719de4db1ddb7553889fb28f4f723aa732023e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/grid_sampler_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 {
+  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 const char* name = "aten::grid_sampler";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "grid_sampler(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);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardshrink_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardshrink_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..645bc0425c92b658e7a8b5e89b7ccb5d060e7c0b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardshrink_backward_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/hardshrink_backward_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_hardshrink_backward_out : public at::meta::structured_hardshrink_backward {
+void impl(const at::Tensor & grad_out, const at::Tensor & self, const at::Scalar & lambd, const at::Tensor & grad_input);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardsigmoid_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardsigmoid_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..48f27a1f4f46ea1ed96ea6ed37f9f41170898b4b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardsigmoid_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 hardsigmoid(const at::Tensor & self);
+TORCH_API at::Tensor & hardsigmoid_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & hardsigmoid_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & hardsigmoid_(at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardswish_backward_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardswish_backward_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..1dfbcb731ffe94c6002a6fbaf9c76fb3af12bd5a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardswish_backward_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & hardswish_backward_out(at::Tensor & out, const at::Tensor & grad_output, const at::Tensor & self);
+TORCH_API at::Tensor & hardswish_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardswish_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardswish_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a52f8a80932e8b70c47e75c5bac73971a385bc7e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardswish_backward_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 hardswish_backward(const at::Tensor & grad_output, const at::Tensor & self);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardtanh_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardtanh_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c4d116ca839f93dbcbca6193c45de26de95c21d8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardtanh_backward_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 hardtanh_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & min_val, const at::Scalar & max_val);
+TORCH_API at::Tensor & hardtanh_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & min_val, const at::Scalar & max_val);
+TORCH_API at::Tensor & hardtanh_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & min_val, const at::Scalar & max_val, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardtanh_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardtanh_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..693598cca66a3e20f27f08c6514e68f4b76b96f1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hardtanh_native.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 hardtanh(const at::Tensor & self, const at::Scalar & min_val=-1, const at::Scalar & max_val=1);
+TORCH_API at::Tensor & hardtanh_out(const at::Tensor & self, const at::Scalar & min_val, const at::Scalar & max_val, at::Tensor & out);
+TORCH_API at::Tensor & hardtanh_(at::Tensor & self, const at::Scalar & min_val=-1, const at::Scalar & max_val=1);
+TORCH_API at::Tensor hardtanh_quantized_cpu(const at::Tensor & self, const at::Scalar & min_val=-1, const at::Scalar & max_val=1);
+TORCH_API at::Tensor & hardtanh_out_quantized_cpu(const at::Tensor & self, const at::Scalar & min_val, const at::Scalar & max_val, at::Tensor & out);
+TORCH_API at::Tensor & hardtanh_quantized_cpu_(at::Tensor & self, const at::Scalar & min_val=-1, const at::Scalar & max_val=1);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/heaviside_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/heaviside_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8830ec9992f72c015bf6c9d66e76cdbf561ba9aa
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/heaviside_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 heaviside(const at::Tensor & self, const at::Tensor & values);
+TORCH_API at::Tensor & heaviside_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & values);
+TORCH_API at::Tensor & heaviside_outf(const at::Tensor & self, const at::Tensor & values, at::Tensor & out);
+TORCH_API at::Tensor & heaviside_(at::Tensor & self, const at::Tensor & values);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/heaviside_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/heaviside_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..eb9ad192e412ea454b7c6ba3d0987fd0563d105c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/heaviside_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 heaviside(const at::Tensor & self, const at::Tensor & values);
+TORCH_API at::Tensor & heaviside_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & values);
+TORCH_API at::Tensor & heaviside_outf(const at::Tensor & self, const at::Tensor & values, at::Tensor & out);
+TORCH_API at::Tensor & heaviside_(at::Tensor & self, const at::Tensor & values);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/heaviside_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/heaviside_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..cae8e7f3c49d558f8c654a65e09b069e90826b82
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/heaviside_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/heaviside_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_heaviside_out : public at::meta::structured_heaviside {
+void impl(const at::Tensor & self, const at::Tensor & values, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hinge_embedding_loss_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hinge_embedding_loss_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..87cc472b31b8208f5920ffb1fc0e04b176458e01
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hinge_embedding_loss_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 hinge_embedding_loss {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, double, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::hinge_embedding_loss";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "hinge_embedding_loss(Tensor self, Tensor target, float margin=1.0, int reduction=Mean) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & target, double margin, int64_t reduction);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & target, double margin, int64_t reduction);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/histogram.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/histogram.h
new file mode 100644
index 0000000000000000000000000000000000000000..e7f2fa37a43bcc23e17f028ae0b0c44b186962b2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/histogram.h
@@ -0,0 +1,59 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/histogram_ops.h>
+
+namespace at {
+
+
+// aten::histogram.bins_tensor_out(Tensor self, Tensor bins, *, Tensor? weight=None, bool density=False, Tensor(a!) hist, Tensor(b!) bin_edges) -> (Tensor(a!) hist, Tensor(b!) bin_edges)
+inline ::std::tuple<at::Tensor &,at::Tensor &> histogram_out(at::Tensor & hist, at::Tensor & bin_edges, const at::Tensor & self, const at::Tensor & bins, const ::std::optional<at::Tensor> & weight={}, bool density=false) {
+    return at::_ops::histogram_bins_tensor_out::call(self, bins, weight, density, hist, bin_edges);
+}
+// aten::histogram.bins_tensor_out(Tensor self, Tensor bins, *, Tensor? weight=None, bool density=False, Tensor(a!) hist, Tensor(b!) bin_edges) -> (Tensor(a!) hist, Tensor(b!) bin_edges)
+inline ::std::tuple<at::Tensor &,at::Tensor &> histogram_outf(const at::Tensor & self, const at::Tensor & bins, const ::std::optional<at::Tensor> & weight, bool density, at::Tensor & hist, at::Tensor & bin_edges) {
+    return at::_ops::histogram_bins_tensor_out::call(self, bins, weight, density, hist, bin_edges);
+}
+
+// aten::histogram.bins_tensor(Tensor self, Tensor bins, *, Tensor? weight=None, bool density=False) -> (Tensor hist, Tensor bin_edges)
+inline ::std::tuple<at::Tensor,at::Tensor> histogram(const at::Tensor & self, const at::Tensor & bins, const ::std::optional<at::Tensor> & weight={}, bool density=false) {
+    return at::_ops::histogram_bins_tensor::call(self, bins, weight, density);
+}
+
+// aten::histogram.bin_ct_out(Tensor self, int bins=100, *, float[]? range=None, Tensor? weight=None, bool density=False, Tensor(a!) hist, Tensor(b!) bin_edges) -> (Tensor(a!) hist, Tensor(b!) bin_edges)
+inline ::std::tuple<at::Tensor &,at::Tensor &> histogram_out(at::Tensor & hist, at::Tensor & bin_edges, const at::Tensor & self, int64_t bins=100, ::std::optional<at::ArrayRef<double>> range=::std::nullopt, const ::std::optional<at::Tensor> & weight={}, bool density=false) {
+    return at::_ops::histogram_bin_ct_out::call(self, bins, range, weight, density, hist, bin_edges);
+}
+// aten::histogram.bin_ct_out(Tensor self, int bins=100, *, float[]? range=None, Tensor? weight=None, bool density=False, Tensor(a!) hist, Tensor(b!) bin_edges) -> (Tensor(a!) hist, Tensor(b!) bin_edges)
+inline ::std::tuple<at::Tensor &,at::Tensor &> histogram_outf(const at::Tensor & self, int64_t bins, ::std::optional<at::ArrayRef<double>> range, const ::std::optional<at::Tensor> & weight, bool density, at::Tensor & hist, at::Tensor & bin_edges) {
+    return at::_ops::histogram_bin_ct_out::call(self, bins, range, weight, density, hist, bin_edges);
+}
+
+// aten::histogram.bin_ct(Tensor self, int bins=100, *, float[]? range=None, Tensor? weight=None, bool density=False) -> (Tensor hist, Tensor bin_edges)
+inline ::std::tuple<at::Tensor,at::Tensor> histogram(const at::Tensor & self, int64_t bins=100, ::std::optional<at::ArrayRef<double>> range=::std::nullopt, const ::std::optional<at::Tensor> & weight={}, bool density=false) {
+    return at::_ops::histogram_bin_ct::call(self, bins, range, weight, density);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/histogramdd_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/histogramdd_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..79f0407cf20d6d9c3128fb83f840d5d6ea5ad136
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/histogramdd_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,::std::vector<at::Tensor>> histogramdd(const at::Tensor & self, at::IntArrayRef bins, ::std::optional<at::ArrayRef<double>> range=::std::nullopt, const ::std::optional<at::Tensor> & weight={}, bool density=false);
+TORCH_API ::std::tuple<at::Tensor,::std::vector<at::Tensor>> histogramdd(const at::Tensor & self, int64_t bins, ::std::optional<at::ArrayRef<double>> range=::std::nullopt, const ::std::optional<at::Tensor> & weight={}, bool density=false);
+TORCH_API ::std::tuple<at::Tensor,::std::vector<at::Tensor>> histogramdd(const at::Tensor & self, at::TensorList bins, ::std::optional<at::ArrayRef<double>> range=::std::nullopt, const ::std::optional<at::Tensor> & weight={}, bool density=false);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hsplit_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hsplit_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..2a209e5d64c4e3afb8afe164dd2f6e2e2dc87292
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hsplit_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 hsplit_int {
+  using schema = ::std::vector<at::Tensor> (const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::hsplit";
+  static constexpr const char* overload_name = "int";
+  static constexpr const char* schema_str = "hsplit.int(Tensor(a -> *) self, int sections) -> Tensor(a)[]";
+  static ::std::vector<at::Tensor> call(const at::Tensor & self, int64_t sections);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t sections);
+};
+
+struct TORCH_API hsplit_array {
+  using schema = ::std::vector<at::Tensor> (const at::Tensor &, at::IntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::hsplit";
+  static constexpr const char* overload_name = "array";
+  static constexpr const char* schema_str = "hsplit.array(Tensor(a -> *) self, int[] indices) -> Tensor(a)[]";
+  static ::std::vector<at::Tensor> call(const at::Tensor & self, at::IntArrayRef indices);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef indices);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hspmm_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hspmm_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..cc51ecf489bd19bcd6b58c57feac6a4ece349a04
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hspmm_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 hspmm_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 const char* name = "aten::hspmm";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "hspmm.out(Tensor mat1, Tensor mat2, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & mat1, const at::Tensor & mat2, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & mat1, const at::Tensor & mat2, at::Tensor & out);
+};
+
+struct TORCH_API hspmm {
+  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 const char* name = "aten::hspmm";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "hspmm(Tensor mat1, Tensor mat2) -> Tensor";
+  static at::Tensor call(const at::Tensor & mat1, const at::Tensor & mat2);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & mat1, const at::Tensor & mat2);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/huber_loss.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/huber_loss.h
new file mode 100644
index 0000000000000000000000000000000000000000..f9afd897f7092f2f66e8052ef7827bec18d6399f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/huber_loss.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/huber_loss_ops.h>
+
+namespace at {
+
+
+// aten::huber_loss.out(Tensor self, Tensor target, int reduction=Mean, float delta=1.0, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & huber_loss_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & target, int64_t reduction=at::Reduction::Mean, double delta=1.0) {
+    return at::_ops::huber_loss_out::call(self, target, reduction, delta, out);
+}
+// aten::huber_loss.out(Tensor self, Tensor target, int reduction=Mean, float delta=1.0, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & huber_loss_outf(const at::Tensor & self, const at::Tensor & target, int64_t reduction, double delta, at::Tensor & out) {
+    return at::_ops::huber_loss_out::call(self, target, reduction, delta, out);
+}
+
+// aten::huber_loss(Tensor self, Tensor target, int reduction=Mean, float delta=1.0) -> Tensor
+inline at::Tensor huber_loss(const at::Tensor & self, const at::Tensor & target, int64_t reduction=at::Reduction::Mean, double delta=1.0) {
+    return at::_ops::huber_loss::call(self, target, reduction, delta);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hypot_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hypot_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..74c8b0be77d519752e890de2cf6615c7b9f35576
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hypot_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 hypot(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & hypot_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & hypot_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & hypot_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hypot_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hypot_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..bbafecd16bf155744c501d5f23ac031e016d30f1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hypot_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_hypot : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & self, const at::Tensor & other);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hypot_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hypot_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..d1c73bd42cacf17bc4b6f11174d190f1ba39bed2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/hypot_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/hypot_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_hypot_out : public at::meta::structured_hypot {
+void impl(const at::Tensor & self, const at::Tensor & other, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/i0_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/i0_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a5b8ef18587d3f8d59d13f0c0fe1295da5d9ad76
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/i0_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 i0(const at::Tensor & self);
+TORCH_API at::Tensor & i0_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/igamma.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/igamma.h
new file mode 100644
index 0000000000000000000000000000000000000000..ad3cd24bb0b7e339b24721fbe94413dae8a5cdd9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/igamma.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/igamma_ops.h>
+
+namespace at {
+
+
+// aten::igamma.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & igamma_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::igamma_out::call(self, other, out);
+}
+// aten::igamma.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & igamma_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out) {
+    return at::_ops::igamma_out::call(self, other, out);
+}
+
+// aten::igamma(Tensor self, Tensor other) -> Tensor
+inline at::Tensor igamma(const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::igamma::call(self, other);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/igamma_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/igamma_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a0f88d60642d8ee907ea208e56f2a5afc64048fe
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/igamma_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 igamma(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & igamma_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & igamma_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & igamma_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/igammac_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/igammac_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..96f230a6e636289fb18ea717a13bc60f972f625a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/igammac_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 igammac(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & igammac_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & igammac_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & igammac_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/im2col_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/im2col_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4f2255e5e4b9d7c19dde5af1e90dab15048e4001
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/im2col_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 im2col(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor & im2col_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor & im2col_outf(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/im2col_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/im2col_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..14bde2b939b7f5ca89ab24add4707dc8ee4d2308
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/im2col_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor im2col_cpu(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor & im2col_out_cpu(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride, at::Tensor & out);
+TORCH_API at::Tensor im2col_cuda(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor & im2col_out_cuda(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/imag_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/imag_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..64c0b3335540600420c890f63b7824daf592de16
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/imag_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 imag {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::imag";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "imag(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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..7504003ea22b7dc4644e9f741718c928b0a994f7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 index(const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices);
+TORCH_API at::Tensor & index_out(at::Tensor & out, const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices);
+TORCH_API at::Tensor & index_outf(const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_fill_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_fill_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8ab5e9eb77031f9066dacf3bed2fed7f53ea459b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_fill_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 index_fill(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value);
+TORCH_API at::Tensor & index_fill_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value);
+TORCH_API at::Tensor & index_fill_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, at::Tensor & out);
+TORCH_API at::Tensor index_fill(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & value);
+TORCH_API at::Tensor & index_fill_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & value);
+TORCH_API at::Tensor & index_fill_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & value, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_put_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_put_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..40adfc47343d6568244bbf7a34db5162f0f5c6cf
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_put_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 index_put(const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate=false);
+TORCH_API at::Tensor & index_put_out(const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate, at::Tensor & out);
+TORCH_API at::Tensor & index_put_(at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate=false);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_put_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_put_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..d07305d002af2c27e41a70427850e2361060b576
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_put_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 index_put_ {
+  using schema = at::Tensor & (at::Tensor &, const c10::List<::std::optional<at::Tensor>> &, const at::Tensor &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::index_put_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "index_put_(Tensor(a!) self, Tensor?[] indices, Tensor values, bool accumulate=False) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate);
+};
+
+struct TORCH_API index_put {
+  using schema = at::Tensor (const at::Tensor &, const c10::List<::std::optional<at::Tensor>> &, const at::Tensor &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::index_put";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "index_put(Tensor self, Tensor?[] indices, Tensor values, bool accumulate=False) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate);
+};
+
+struct TORCH_API index_put_out {
+  using schema = at::Tensor & (const at::Tensor &, const c10::List<::std::optional<at::Tensor>> &, const at::Tensor &, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::index_put";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "index_put.out(Tensor self, Tensor?[] indices, Tensor values, bool accumulate=False, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, const at::Tensor & values, bool accumulate, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_reduce_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_reduce_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..069c142d51c2823842c83b2c4c529721f62ecd9a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_reduce_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 index_reduce(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & source, c10::string_view reduce, bool include_self=true);
+TORCH_API at::Tensor & index_reduce_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & source, c10::string_view reduce, bool include_self=true);
+TORCH_API at::Tensor & index_reduce_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & source, c10::string_view reduce, bool include_self, at::Tensor & out);
+TORCH_API at::Tensor & index_reduce_(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & source, c10::string_view reduce, bool include_self=true);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_select_backward_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_select_backward_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..766d4e11f2a8fc2688f27a8b272b6ebce61495fc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_select_backward_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 index_select_backward(const at::Tensor & grad, at::IntArrayRef self_sizes, int64_t dim, const at::Tensor & index);
+TORCH_API at::Tensor index_select_backward_symint(const at::Tensor & grad, c10::SymIntArrayRef self_sizes, int64_t dim, const at::Tensor & index);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_select_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_select_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..0344fb0328be55999266ca9b9e0833133403de77
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/index_select_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 index_select(const at::Tensor & self, at::Dimname dim, const at::Tensor & index);
+TORCH_API at::Tensor & index_select_out(at::Tensor & out, const at::Tensor & self, at::Dimname dim, const at::Tensor & index);
+TORCH_API at::Tensor & index_select_outf(const at::Tensor & self, at::Dimname dim, const at::Tensor & index, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/indices_copy_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/indices_copy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..9a4c6ef371be3b5155a9614510f2955106ca0add
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/indices_copy_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & indices_copy_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor indices_copy(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/inner_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/inner_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..3916ecdf230975bce85134c8ff4b9aea17a0673d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/inner_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 inner {
+  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 const char* name = "aten::inner";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "inner(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 inner_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 const char* name = "aten::inner";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "inner.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/inverse_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/inverse_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..0c2790b1388072cc79512fd3a2fc204963f3226d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/inverse_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 inverse(const at::Tensor & self);
+TORCH_API at::Tensor & inverse_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & inverse_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_conj_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_conj_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..11893f51d46b9c84cf31d56f4d36a04235501e24
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_conj_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API bool is_conj(const at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_distributed_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_distributed_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..aa77604aa781f53ce3e872d9ac477d40d6fd3b30
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_distributed_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API is_distributed {
+  using schema = bool (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::is_distributed";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "is_distributed(Tensor self) -> bool";
+  static bool call(const at::Tensor & self);
+  static bool redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_inference.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_inference.h
new file mode 100644
index 0000000000000000000000000000000000000000..d1f8d70b40e8aadb64ac4e75739a99b4d7118dbf
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_inference.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/is_inference_ops.h>
+
+namespace at {
+
+
+// aten::is_inference(Tensor self) -> bool
+inline bool __dispatch_is_inference(const at::Tensor & self) {
+    return at::_ops::is_inference::call(self);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_nonzero_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_nonzero_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5528b777646016b5895f809c1089989b2cab85ac
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_nonzero_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API bool is_nonzero(const at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_set_to_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_set_to_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..2ebc1fd377d8390669468304b39b724d34d9d4eb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/is_set_to_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API bool is_set_to(const at::Tensor & self, const at::Tensor & tensor);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isfinite_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isfinite_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..2d10e9df927136023b3e6142e13c653454c21184
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isfinite_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 isfinite(const at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isinf_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isinf_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..976dc2c12b4f0afcf5bd519d03a069b7ecd28b4d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isinf_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 isinf {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::isinf";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "isinf(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 isinf_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 const char* name = "aten::isinf";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "isinf.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isnan_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isnan_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..fe47bb7b051ff1bd85ae8b4e6e3b59837d478e2d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isnan_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 isnan(const at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isneginf_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isneginf_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c15158f09213113bed4a3d760ec562413210f4b2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isneginf_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 isneginf(const at::Tensor & self);
+TORCH_API at::Tensor & isneginf_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & isneginf_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isposinf_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isposinf_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..574f7158b66f9c244118ccb0cef183169fb20878
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isposinf_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 isposinf(const at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isposinf_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isposinf_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..dd96c61eed10f4d797da603efc2566e4500966ce
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isposinf_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 isposinf(const at::Tensor & self);
+TORCH_API at::Tensor & isposinf_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & isposinf_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isposinf_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isposinf_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d867a75eae2f636e414a43a5b317c3ebca51d7d2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isposinf_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 isposinf(const at::Tensor & self);
+TORCH_API at::Tensor & isposinf_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & isposinf_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isreal_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isreal_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..48d577fb823fb827a9b084648d7d75875fb0d674
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/isreal_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 isreal(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/istft_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/istft_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..509cf2a072cc56b7b2ad4d8563912e72da0ecfa5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/istft_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 istft {
+  using schema = at::Tensor (const at::Tensor &, int64_t, ::std::optional<int64_t>, ::std::optional<int64_t>, const ::std::optional<at::Tensor> &, bool, bool, ::std::optional<bool>, ::std::optional<int64_t>, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::istft";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "istft(Tensor self, int n_fft, int? hop_length=None, int? win_length=None, Tensor? window=None, bool center=True, bool normalized=False, bool? onesided=None, int? length=None, bool return_complex=False) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t n_fft, ::std::optional<int64_t> hop_length, ::std::optional<int64_t> win_length, const ::std::optional<at::Tensor> & window, bool center, bool normalized, ::std::optional<bool> onesided, ::std::optional<int64_t> length, bool return_complex);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t n_fft, ::std::optional<int64_t> hop_length, ::std::optional<int64_t> win_length, const ::std::optional<at::Tensor> & window, bool center, bool normalized, ::std::optional<bool> onesided, ::std::optional<int64_t> length, bool return_complex);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/kl_div_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/kl_div_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..845c736e4143637c975cc0d283d19f1ebb20288f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/kl_div_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 kl_div(const at::Tensor & self, const at::Tensor & target, int64_t reduction=at::Reduction::Mean, bool log_target=false);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/kthvalue.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/kthvalue.h
new file mode 100644
index 0000000000000000000000000000000000000000..90137245f06ab39c4a862c83090e0be458827c1f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/kthvalue.h
@@ -0,0 +1,163 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/kthvalue_ops.h>
+
+namespace at {
+
+
+// aten::kthvalue(Tensor self, SymInt k, int dim=-1, bool keepdim=False) -> (Tensor values, Tensor indices)
+inline ::std::tuple<at::Tensor,at::Tensor> kthvalue(const at::Tensor & self, int64_t k, int64_t dim=-1, bool keepdim=false) {
+    return at::_ops::kthvalue::call(self, k, dim, keepdim);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor,at::Tensor> kthvalue(const at::Tensor & self, int64_t k, int64_t dim=-1, bool keepdim=false) {
+    return at::_ops::kthvalue::call(self, k, dim, keepdim);
+  }
+}
+
+// aten::kthvalue(Tensor self, SymInt k, int dim=-1, bool keepdim=False) -> (Tensor values, Tensor indices)
+inline ::std::tuple<at::Tensor,at::Tensor> kthvalue_symint(const at::Tensor & self, c10::SymInt k, int64_t dim=-1, bool keepdim=false) {
+    return at::_ops::kthvalue::call(self, k, dim, keepdim);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor,at::Tensor> kthvalue(const at::Tensor & self, c10::SymInt k, int64_t dim=-1, bool keepdim=false) {
+    return at::_ops::kthvalue::call(self, k, dim, keepdim);
+  }
+}
+
+// aten::kthvalue.values(Tensor self, SymInt k, int dim=-1, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, int64_t k, int64_t dim=-1, bool keepdim=false) {
+    return at::_ops::kthvalue_values::call(self, k, dim, keepdim, values, indices);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, int64_t k, int64_t dim=-1, bool keepdim=false) {
+    return at::_ops::kthvalue_values::call(self, k, dim, keepdim, values, indices);
+  }
+}
+
+// aten::kthvalue.values(Tensor self, SymInt k, int dim=-1, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_outf(const at::Tensor & self, int64_t k, int64_t dim, bool keepdim, at::Tensor & values, at::Tensor & indices) {
+    return at::_ops::kthvalue_values::call(self, k, dim, keepdim, values, indices);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_outf(const at::Tensor & self, int64_t k, int64_t dim, bool keepdim, at::Tensor & values, at::Tensor & indices) {
+    return at::_ops::kthvalue_values::call(self, k, dim, keepdim, values, indices);
+  }
+}
+
+// aten::kthvalue.values(Tensor self, SymInt k, int dim=-1, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_symint_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, c10::SymInt k, int64_t dim=-1, bool keepdim=false) {
+    return at::_ops::kthvalue_values::call(self, k, dim, keepdim, values, indices);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, c10::SymInt k, int64_t dim=-1, bool keepdim=false) {
+    return at::_ops::kthvalue_values::call(self, k, dim, keepdim, values, indices);
+  }
+}
+
+// aten::kthvalue.values(Tensor self, SymInt k, int dim=-1, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_symint_outf(const at::Tensor & self, c10::SymInt k, int64_t dim, bool keepdim, at::Tensor & values, at::Tensor & indices) {
+    return at::_ops::kthvalue_values::call(self, k, dim, keepdim, values, indices);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_outf(const at::Tensor & self, c10::SymInt k, int64_t dim, bool keepdim, at::Tensor & values, at::Tensor & indices) {
+    return at::_ops::kthvalue_values::call(self, k, dim, keepdim, values, indices);
+  }
+}
+
+// aten::kthvalue.dimname(Tensor self, SymInt k, Dimname dim, bool keepdim=False) -> (Tensor values, Tensor indices)
+inline ::std::tuple<at::Tensor,at::Tensor> kthvalue(const at::Tensor & self, int64_t k, at::Dimname dim, bool keepdim=false) {
+    return at::_ops::kthvalue_dimname::call(self, k, dim, keepdim);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor,at::Tensor> kthvalue(const at::Tensor & self, int64_t k, at::Dimname dim, bool keepdim=false) {
+    return at::_ops::kthvalue_dimname::call(self, k, dim, keepdim);
+  }
+}
+
+// aten::kthvalue.dimname(Tensor self, SymInt k, Dimname dim, bool keepdim=False) -> (Tensor values, Tensor indices)
+inline ::std::tuple<at::Tensor,at::Tensor> kthvalue_symint(const at::Tensor & self, c10::SymInt k, at::Dimname dim, bool keepdim=false) {
+    return at::_ops::kthvalue_dimname::call(self, k, dim, keepdim);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor,at::Tensor> kthvalue(const at::Tensor & self, c10::SymInt k, at::Dimname dim, bool keepdim=false) {
+    return at::_ops::kthvalue_dimname::call(self, k, dim, keepdim);
+  }
+}
+
+// aten::kthvalue.dimname_out(Tensor self, SymInt k, Dimname dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, int64_t k, at::Dimname dim, bool keepdim=false) {
+    return at::_ops::kthvalue_dimname_out::call(self, k, dim, keepdim, values, indices);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, int64_t k, at::Dimname dim, bool keepdim=false) {
+    return at::_ops::kthvalue_dimname_out::call(self, k, dim, keepdim, values, indices);
+  }
+}
+
+// aten::kthvalue.dimname_out(Tensor self, SymInt k, Dimname dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_outf(const at::Tensor & self, int64_t k, at::Dimname dim, bool keepdim, at::Tensor & values, at::Tensor & indices) {
+    return at::_ops::kthvalue_dimname_out::call(self, k, dim, keepdim, values, indices);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_outf(const at::Tensor & self, int64_t k, at::Dimname dim, bool keepdim, at::Tensor & values, at::Tensor & indices) {
+    return at::_ops::kthvalue_dimname_out::call(self, k, dim, keepdim, values, indices);
+  }
+}
+
+// aten::kthvalue.dimname_out(Tensor self, SymInt k, Dimname dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_symint_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, c10::SymInt k, at::Dimname dim, bool keepdim=false) {
+    return at::_ops::kthvalue_dimname_out::call(self, k, dim, keepdim, values, indices);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, c10::SymInt k, at::Dimname dim, bool keepdim=false) {
+    return at::_ops::kthvalue_dimname_out::call(self, k, dim, keepdim, values, indices);
+  }
+}
+
+// aten::kthvalue.dimname_out(Tensor self, SymInt k, Dimname dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_symint_outf(const at::Tensor & self, c10::SymInt k, at::Dimname dim, bool keepdim, at::Tensor & values, at::Tensor & indices) {
+    return at::_ops::kthvalue_dimname_out::call(self, k, dim, keepdim, values, indices);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor &,at::Tensor &> kthvalue_outf(const at::Tensor & self, c10::SymInt k, at::Dimname dim, bool keepdim, at::Tensor & values, at::Tensor & indices) {
+    return at::_ops::kthvalue_dimname_out::call(self, k, dim, keepdim, values, indices);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/l1_loss.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/l1_loss.h
new file mode 100644
index 0000000000000000000000000000000000000000..bd1d7cf35f101ec8cf8852fcc9ca853841edfb3c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/l1_loss.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/l1_loss_ops.h>
+
+namespace at {
+
+
+// aten::l1_loss(Tensor self, Tensor target, int reduction=Mean) -> Tensor
+inline at::Tensor l1_loss(const at::Tensor & self, const at::Tensor & target, int64_t reduction=at::Reduction::Mean) {
+    return at::_ops::l1_loss::call(self, target, reduction);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/layer_norm_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/layer_norm_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..d61446faaf55cad14ca0c798a8e581c3f236ae55
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/layer_norm_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 layer_norm {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, double, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::layer_norm";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "layer_norm(Tensor input, SymInt[] normalized_shape, Tensor? weight=None, Tensor? bias=None, float eps=1e-05, bool cudnn_enable=True) -> Tensor";
+  static at::Tensor call(const at::Tensor & input, c10::SymIntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps, bool cudnn_enable);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, c10::SymIntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps, bool cudnn_enable);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lcm_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lcm_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..1bf06e75578da55bbedc3a5f83b6d649bef288d6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lcm_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/lcm_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_lcm_out : public at::meta::structured_lcm {
+void impl(const at::Tensor & self, const at::Tensor & other, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ldexp.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ldexp.h
new file mode 100644
index 0000000000000000000000000000000000000000..468aeb29878f358da7b56ea8725cd67697ceb208
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ldexp.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/ldexp_ops.h>
+
+namespace at {
+
+
+// aten::ldexp.Tensor(Tensor self, Tensor other) -> Tensor
+inline at::Tensor ldexp(const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::ldexp_Tensor::call(self, other);
+}
+
+// aten::ldexp_(Tensor(a!) self, Tensor other) -> Tensor(a!)
+inline at::Tensor & ldexp_(at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::ldexp_::call(self, other);
+}
+
+// aten::ldexp.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & ldexp_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::ldexp_out::call(self, other, out);
+}
+// aten::ldexp.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & ldexp_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out) {
+    return at::_ops::ldexp_out::call(self, other, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/le_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/le_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6a2d604df85df4dfdea2c169a76a7d067dd0a18e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/le_cuda_dispatch.h
@@ -0,0 +1,35 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 le(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & le_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & le_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & le_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor le(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & le_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & le_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & le_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..48422eac537f670fbd6dfcc267862c745f2464fa
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 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 cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..0f0d6e14ccc384357cbb078dbcd38f5339179da8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_leaky_relu : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & self, const at::Scalar & negative_slope);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5f40854e135af193a7ed0560f787fde385956344
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 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 meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..e84b68ba872bc1251751df0331b116710fe0448b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/leaky_relu_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 leaky_relu_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 const char* name = "aten::leaky_relu";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "leaky_relu.out(Tensor self, Scalar negative_slope=0.01, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Scalar & negative_slope, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & negative_slope, at::Tensor & out);
+};
+
+struct TORCH_API leaky_relu {
+  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 const char* name = "aten::leaky_relu";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "leaky_relu(Tensor self, Scalar negative_slope=0.01) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Scalar & negative_slope);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & negative_slope);
+};
+
+struct TORCH_API leaky_relu_ {
+  using schema = at::Tensor & (at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::leaky_relu_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "leaky_relu_(Tensor(a!) self, Scalar negative_slope=0.01) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Scalar & negative_slope);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Scalar & negative_slope);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lerp_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lerp_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..69d5dffc337a3bbabcc2676abf0189a290e50283
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lerp_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 lerp(const at::Tensor & self, const at::Tensor & end, const at::Scalar & weight);
+TORCH_API at::Tensor & lerp_(at::Tensor & self, const at::Tensor & end, const at::Scalar & weight);
+TORCH_API at::Tensor lerp(const at::Tensor & self, const at::Tensor & end, const at::Tensor & weight);
+TORCH_API at::Tensor & lerp_(at::Tensor & self, const at::Tensor & end, const at::Tensor & weight);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lgamma_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lgamma_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..65cb8ccb82a5e5fdb096790bebf61c21652bd202
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lgamma_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 lgamma(const at::Tensor & self);
+TORCH_API at::Tensor & lgamma_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lgamma_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lgamma_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..18b5477afcb17e348d87cf52f589cd6715639a91
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lgamma_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 lgamma(const at::Tensor & self);
+TORCH_API at::Tensor & lgamma_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & lgamma_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & lgamma_(at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lgamma_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lgamma_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..5b7967e4ed4d2c21808b00f6f3fe60d7a101dd0d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lgamma_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 lgamma_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 const char* name = "aten::lgamma";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "lgamma.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);
+};
+
+struct TORCH_API lgamma_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::lgamma_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "lgamma_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API lgamma {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::lgamma";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "lgamma(Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lift_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lift_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..10c8845ea70392b8e083a16b6a99e90df88c4824
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lift_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 lift(const at::Tensor & self);
+TORCH_API at::Tensor & lift_out(const at::Tensor & self, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg__powsum_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg__powsum_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b327dcdfe878e81fe76bc96798785a088caec938
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg__powsum_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 linalg__powsum(const at::Tensor & self, const at::Scalar & ord=2, at::OptionalIntArrayRef dim=::std::nullopt, bool keepdim=false, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg__powsum_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg__powsum_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..618c38163f9673d19c488c77d9f4a4c9d06144c1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg__powsum_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API linalg__powsum {
+  using schema = at::Tensor (const at::Tensor &, const at::Scalar &, at::OptionalIntArrayRef, bool, ::std::optional<at::ScalarType>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg__powsum";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "linalg__powsum(Tensor self, Scalar ord=2, int[1]? dim=None, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Scalar & ord, at::OptionalIntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & ord, at::OptionalIntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_cholesky_ex_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_cholesky_ex_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..258e83e9280e0116f81d5d5071c784706c588331
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_cholesky_ex_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_linalg_cholesky_ex : public at::impl::MetaBase {
+
+
+    void meta(const at::Tensor & self, bool upper, bool check_errors);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_cholesky_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_cholesky_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..4cc42cfed6298a51f564de6af208ccfe265800f4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_cholesky_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_diagonal.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_diagonal.h
new file mode 100644
index 0000000000000000000000000000000000000000..552a37987ff48d35ec6fcbf2f488ac6affcf2e17
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_diagonal.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/linalg_diagonal_ops.h>
+
+namespace at {
+
+
+// aten::linalg_diagonal(Tensor(a) A, *, int offset=0, int dim1=-2, int dim2=-1) -> Tensor(a)
+inline at::Tensor linalg_diagonal(const at::Tensor & A, int64_t offset=0, int64_t dim1=-2, int64_t dim2=-1) {
+    return at::_ops::linalg_diagonal::call(A, offset, dim1, dim2);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigh.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigh.h
new file mode 100644
index 0000000000000000000000000000000000000000..ba5bcb116dd51221203fcc67e73b135aa83768e5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigh.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/linalg_eigh_ops.h>
+
+namespace at {
+
+
+// aten::linalg_eigh(Tensor self, str UPLO="L") -> (Tensor eigenvalues, Tensor eigenvectors)
+inline ::std::tuple<at::Tensor,at::Tensor> linalg_eigh(const at::Tensor & self, c10::string_view UPLO="L") {
+    return at::_ops::linalg_eigh::call(self, UPLO);
+}
+
+// aten::linalg_eigh.eigvals(Tensor self, str UPLO="L", *, Tensor(a!) eigvals, Tensor(b!) eigvecs) -> (Tensor(a!) eigenvalues, Tensor(b!) eigenvectors)
+inline ::std::tuple<at::Tensor &,at::Tensor &> linalg_eigh_out(at::Tensor & eigvals, at::Tensor & eigvecs, const at::Tensor & self, c10::string_view UPLO="L") {
+    return at::_ops::linalg_eigh_eigvals::call(self, UPLO, eigvals, eigvecs);
+}
+// aten::linalg_eigh.eigvals(Tensor self, str UPLO="L", *, Tensor(a!) eigvals, Tensor(b!) eigvecs) -> (Tensor(a!) eigenvalues, Tensor(b!) eigenvectors)
+inline ::std::tuple<at::Tensor &,at::Tensor &> linalg_eigh_outf(const at::Tensor & self, c10::string_view UPLO, at::Tensor & eigvals, at::Tensor & eigvecs) {
+    return at::_ops::linalg_eigh_eigvals::call(self, UPLO, eigvals, eigvecs);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigh_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigh_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4020670263c9f777feebda2ecad8f2dcb7b335d5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigh_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,at::Tensor> linalg_eigh(const at::Tensor & self, c10::string_view UPLO="L");
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> linalg_eigh_out(at::Tensor & eigvals, at::Tensor & eigvecs, const at::Tensor & self, c10::string_view UPLO="L");
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> linalg_eigh_outf(const at::Tensor & self, c10::string_view UPLO, at::Tensor & eigvals, at::Tensor & eigvecs);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigvals_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigvals_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..e6f2d4310e59d4109ef03d84be607beaf0febbaa
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigvals_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_eigvals(const at::Tensor & self);
+TORCH_API at::Tensor & linalg_eigvals_out(const at::Tensor & self, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigvalsh_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigvalsh_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..42f0eec32cc85002097cc98603d0ede3e5d3506f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_eigvalsh_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_eigvalsh(const at::Tensor & self, c10::string_view UPLO="L");
+TORCH_API at::Tensor & linalg_eigvalsh_out(const at::Tensor & self, c10::string_view UPLO, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_inv_ex_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_inv_ex_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..aee282cd54f7ae6baba77e4199c3b6acfda7790d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_inv_ex_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/linalg_inv_ex_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_linalg_inv_ex_out : public at::meta::structured_linalg_inv_ex {
+void impl(const at::Tensor & A, bool check_errors, const at::Tensor & inverse, const at::Tensor & info);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_inv_ex_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_inv_ex_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..9e914e303e9270c361217672376cbeb79b39cd64
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_inv_ex_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API linalg_inv_ex {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_inv_ex";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "linalg_inv_ex(Tensor A, *, bool check_errors=False) -> (Tensor inverse, Tensor info)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & A, bool check_errors);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & A, bool check_errors);
+};
+
+struct TORCH_API linalg_inv_ex_inverse {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, bool, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_inv_ex";
+  static constexpr const char* overload_name = "inverse";
+  static constexpr const char* schema_str = "linalg_inv_ex.inverse(Tensor A, *, bool check_errors=False, Tensor(a!) inverse, Tensor(b!) info) -> (Tensor(a!) inverse, Tensor(b!) info)";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & A, bool check_errors, at::Tensor & inverse, at::Tensor & info);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & A, bool check_errors, at::Tensor & inverse, at::Tensor & info);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ex_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ex_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3f18a88893a2620820cda68b093a7adf4d2559bb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ex_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,at::Tensor,at::Tensor> linalg_ldl_factor_ex(const at::Tensor & self, bool hermitian=false, bool check_errors=false);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ex_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ex_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e5247aa1f2dfa3fe3fae1d29cc8d6cb3332bb56d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ex_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> linalg_ldl_factor_ex(const at::Tensor & self, bool hermitian=false, bool check_errors=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> linalg_ldl_factor_ex_out(at::Tensor & LD, at::Tensor & pivots, at::Tensor & info, const at::Tensor & self, bool hermitian=false, bool check_errors=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> linalg_ldl_factor_ex_outf(const at::Tensor & self, bool hermitian, bool check_errors, at::Tensor & LD, at::Tensor & pivots, at::Tensor & info);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ex_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ex_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..4f30da78251eba051419ab0ee1761e634e64174c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ex_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/linalg_ldl_factor_ex_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_linalg_ldl_factor_ex_out : public at::meta::structured_linalg_ldl_factor_ex {
+void impl(const at::Tensor & self, bool hermitian, bool check_errors, const at::Tensor & LD, const at::Tensor & pivots, const at::Tensor & info);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8fff7845f694aea4cf8cd8162ff8f67f1d7e69c4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_factor_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API linalg_ldl_factor {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_ldl_factor";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "linalg_ldl_factor(Tensor self, *, bool hermitian=False) -> (Tensor LD, Tensor pivots)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, bool hermitian);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, bool hermitian);
+};
+
+struct TORCH_API linalg_ldl_factor_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, bool, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_ldl_factor";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "linalg_ldl_factor.out(Tensor self, *, bool hermitian=False, Tensor(a!) LD, Tensor(b!) pivots) -> (Tensor(a!) LD, Tensor(b!) pivots)";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, bool hermitian, at::Tensor & LD, at::Tensor & pivots);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, bool hermitian, at::Tensor & LD, at::Tensor & pivots);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_solve_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_solve_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a30f2ee3ae110e9f659b0dd31552f37165f2390e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_ldl_solve_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 linalg_ldl_solve(const at::Tensor & LD, const at::Tensor & pivots, const at::Tensor & B, bool hermitian=false);
+TORCH_API at::Tensor & linalg_ldl_solve_out(at::Tensor & out, const at::Tensor & LD, const at::Tensor & pivots, const at::Tensor & B, bool hermitian=false);
+TORCH_API at::Tensor & linalg_ldl_solve_outf(const at::Tensor & LD, const at::Tensor & pivots, const at::Tensor & B, bool hermitian, at::Tensor & out);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lstsq.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lstsq.h
new file mode 100644
index 0000000000000000000000000000000000000000..830d7bf13ab31da173716288eab35e010b15a1fa
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lstsq.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/linalg_lstsq_ops.h>
+
+namespace at {
+
+
+// aten::linalg_lstsq(Tensor self, Tensor b, float? rcond=None, *, str? driver=None) -> (Tensor solution, Tensor residuals, Tensor rank, Tensor singular_values)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> linalg_lstsq(const at::Tensor & self, const at::Tensor & b, ::std::optional<double> rcond=::std::nullopt, ::std::optional<c10::string_view> driver=::std::nullopt) {
+    return at::_ops::linalg_lstsq::call(self, b, rcond, driver);
+}
+
+// aten::linalg_lstsq.out(Tensor self, Tensor b, float? rcond=None, *, str? driver=None, Tensor(a!) solution, Tensor(b!) residuals, Tensor(c!) rank, Tensor(d!) singular_values) -> (Tensor(a!) solution, Tensor(b!) residuals, Tensor(c!) rank, Tensor(d!) singular_values)
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> linalg_lstsq_out(at::Tensor & solution, at::Tensor & residuals, at::Tensor & rank, at::Tensor & singular_values, const at::Tensor & self, const at::Tensor & b, ::std::optional<double> rcond=::std::nullopt, ::std::optional<c10::string_view> driver=::std::nullopt) {
+    return at::_ops::linalg_lstsq_out::call(self, b, rcond, driver, solution, residuals, rank, singular_values);
+}
+// aten::linalg_lstsq.out(Tensor self, Tensor b, float? rcond=None, *, str? driver=None, Tensor(a!) solution, Tensor(b!) residuals, Tensor(c!) rank, Tensor(d!) singular_values) -> (Tensor(a!) solution, Tensor(b!) residuals, Tensor(c!) rank, Tensor(d!) singular_values)
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> linalg_lstsq_outf(const at::Tensor & self, const at::Tensor & b, ::std::optional<double> rcond, ::std::optional<c10::string_view> driver, at::Tensor & solution, at::Tensor & residuals, at::Tensor & rank, at::Tensor & singular_values) {
+    return at::_ops::linalg_lstsq_out::call(self, b, rcond, driver, solution, residuals, rank, singular_values);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f6b57231823da5054c78a795676f4bad754aec76
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,at::Tensor,at::Tensor> linalg_lu(const at::Tensor & A, bool pivot=true);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_factor.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_factor.h
new file mode 100644
index 0000000000000000000000000000000000000000..b4da9e5ded69d529437148db43aa98dbe715fb86
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_factor.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/linalg_lu_factor_ops.h>
+
+namespace at {
+
+
+// aten::linalg_lu_factor(Tensor A, *, bool pivot=True) -> (Tensor LU, Tensor pivots)
+inline ::std::tuple<at::Tensor,at::Tensor> linalg_lu_factor(const at::Tensor & A, bool pivot=true) {
+    return at::_ops::linalg_lu_factor::call(A, pivot);
+}
+
+// aten::linalg_lu_factor.out(Tensor A, *, bool pivot=True, Tensor(a!) LU, Tensor(b!) pivots) -> (Tensor(a!) LU, Tensor(b!) pivots)
+inline ::std::tuple<at::Tensor &,at::Tensor &> linalg_lu_factor_out(at::Tensor & LU, at::Tensor & pivots, const at::Tensor & A, bool pivot=true) {
+    return at::_ops::linalg_lu_factor_out::call(A, pivot, LU, pivots);
+}
+// aten::linalg_lu_factor.out(Tensor A, *, bool pivot=True, Tensor(a!) LU, Tensor(b!) pivots) -> (Tensor(a!) LU, Tensor(b!) pivots)
+inline ::std::tuple<at::Tensor &,at::Tensor &> linalg_lu_factor_outf(const at::Tensor & A, bool pivot, at::Tensor & LU, at::Tensor & pivots) {
+    return at::_ops::linalg_lu_factor_out::call(A, pivot, LU, pivots);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_factor_ex_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_factor_ex_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..0f509939618e5503366569b32f0971e1fab8680c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_factor_ex_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> linalg_lu_factor_ex(const at::Tensor & A, bool pivot=true, bool check_errors=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> linalg_lu_factor_ex_out(at::Tensor & LU, at::Tensor & pivots, at::Tensor & info, const at::Tensor & A, bool pivot=true, bool check_errors=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> linalg_lu_factor_ex_outf(const at::Tensor & A, bool pivot, bool check_errors, at::Tensor & LU, at::Tensor & pivots, at::Tensor & info);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_solve_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_solve_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9554a464fe5094788f2c8bcca57bc5cc0c5ee1b3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_lu_solve_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 linalg_lu_solve(const at::Tensor & LU, const at::Tensor & pivots, const at::Tensor & B, bool left=true, bool adjoint=false);
+TORCH_API at::Tensor & linalg_lu_solve_out(at::Tensor & out, const at::Tensor & LU, const at::Tensor & pivots, const at::Tensor & B, bool left=true, bool adjoint=false);
+TORCH_API at::Tensor & linalg_lu_solve_outf(const at::Tensor & LU, const at::Tensor & pivots, const at::Tensor & B, bool left, bool adjoint, at::Tensor & out);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_matrix_exp_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_matrix_exp_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8dc3539fb64a1ab8f44230dc9ad58b78efdee6bc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_matrix_exp_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 linalg_matrix_exp(const at::Tensor & self);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_matrix_exp_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_matrix_exp_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..dc562d898c9bef25c41b7241368a0c46bc226c4e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_matrix_exp_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor & linalg_matrix_exp_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor linalg_matrix_exp(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_matrix_norm_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_matrix_norm_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..9ca20dc51b3f2da21f4438cb189735a238e11670
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_matrix_norm_ops.h
@@ -0,0 +1,67 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API linalg_matrix_norm {
+  using schema = at::Tensor (const at::Tensor &, const at::Scalar &, at::IntArrayRef, bool, ::std::optional<at::ScalarType>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_matrix_norm";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "linalg_matrix_norm(Tensor self, Scalar ord, int[] dim=[-2,-1], bool keepdim=False, *, ScalarType? dtype=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Scalar & ord, at::IntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & ord, at::IntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+};
+
+struct TORCH_API linalg_matrix_norm_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Scalar &, at::IntArrayRef, bool, ::std::optional<at::ScalarType>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_matrix_norm";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "linalg_matrix_norm.out(Tensor self, Scalar ord, int[] dim=[-2,-1], bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Scalar & ord, at::IntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & ord, at::IntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+};
+
+struct TORCH_API linalg_matrix_norm_str_ord {
+  using schema = at::Tensor (const at::Tensor &, c10::string_view, at::IntArrayRef, bool, ::std::optional<at::ScalarType>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_matrix_norm";
+  static constexpr const char* overload_name = "str_ord";
+  static constexpr const char* schema_str = "linalg_matrix_norm.str_ord(Tensor self, str ord='fro', int[] dim=[-2,-1], bool keepdim=False, *, ScalarType? dtype=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::string_view ord, at::IntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::string_view ord, at::IntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+};
+
+struct TORCH_API linalg_matrix_norm_str_ord_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::string_view, at::IntArrayRef, bool, ::std::optional<at::ScalarType>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_matrix_norm";
+  static constexpr const char* overload_name = "str_ord_out";
+  static constexpr const char* schema_str = "linalg_matrix_norm.str_ord_out(Tensor self, str ord='fro', int[] dim=[-2,-1], bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::string_view ord, at::IntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::string_view ord, at::IntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_norm.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_norm.h
new file mode 100644
index 0000000000000000000000000000000000000000..b3f7e96c67711b4631c42edf1b82b8c130ff57f5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_norm.h
@@ -0,0 +1,59 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/linalg_norm_ops.h>
+
+namespace at {
+
+
+// aten::linalg_norm(Tensor self, Scalar? ord=None, int[1]? dim=None, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor
+inline at::Tensor linalg_norm(const at::Tensor & self, const ::std::optional<at::Scalar> & ord=::std::nullopt, at::OptionalIntArrayRef dim=::std::nullopt, bool keepdim=false, ::std::optional<at::ScalarType> dtype=::std::nullopt) {
+    return at::_ops::linalg_norm::call(self, ord, dim, keepdim, dtype);
+}
+
+// aten::linalg_norm.ord_str(Tensor self, str ord, int[1]? dim=None, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor
+inline at::Tensor linalg_norm(const at::Tensor & self, c10::string_view ord, at::OptionalIntArrayRef dim=::std::nullopt, bool keepdim=false, ::std::optional<at::ScalarType> dtype=::std::nullopt) {
+    return at::_ops::linalg_norm_ord_str::call(self, ord, dim, keepdim, dtype);
+}
+
+// aten::linalg_norm.out(Tensor self, Scalar? ord=None, int[1]? dim=None, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_norm_out(at::Tensor & out, const at::Tensor & self, const ::std::optional<at::Scalar> & ord=::std::nullopt, at::OptionalIntArrayRef dim=::std::nullopt, bool keepdim=false, ::std::optional<at::ScalarType> dtype=::std::nullopt) {
+    return at::_ops::linalg_norm_out::call(self, ord, dim, keepdim, dtype, out);
+}
+// aten::linalg_norm.out(Tensor self, Scalar? ord=None, int[1]? dim=None, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_norm_outf(const at::Tensor & self, const ::std::optional<at::Scalar> & ord, at::OptionalIntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out) {
+    return at::_ops::linalg_norm_out::call(self, ord, dim, keepdim, dtype, out);
+}
+
+// aten::linalg_norm.ord_str_out(Tensor self, str ord, int[1]? dim=None, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_norm_out(at::Tensor & out, const at::Tensor & self, c10::string_view ord, at::OptionalIntArrayRef dim=::std::nullopt, bool keepdim=false, ::std::optional<at::ScalarType> dtype=::std::nullopt) {
+    return at::_ops::linalg_norm_ord_str_out::call(self, ord, dim, keepdim, dtype, out);
+}
+// aten::linalg_norm.ord_str_out(Tensor self, str ord, int[1]? dim=None, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_norm_outf(const at::Tensor & self, c10::string_view ord, at::OptionalIntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out) {
+    return at::_ops::linalg_norm_ord_str_out::call(self, ord, dim, keepdim, dtype, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_pinv_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_pinv_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..f75666228a317e0ce1e78a8fffcee6ddee258e36
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_pinv_native.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_pinv_out(const at::Tensor & self, const ::std::optional<at::Tensor> & atol, const ::std::optional<at::Tensor> & rtol, bool hermitian, at::Tensor & out);
+TORCH_API at::Tensor linalg_pinv(const at::Tensor & self, const ::std::optional<at::Tensor> & atol={}, const ::std::optional<at::Tensor> & rtol={}, bool hermitian=false);
+TORCH_API at::Tensor linalg_pinv(const at::Tensor & self, ::std::optional<double> atol=::std::nullopt, ::std::optional<double> rtol=::std::nullopt, bool hermitian=false);
+TORCH_API at::Tensor & linalg_pinv_out(const at::Tensor & self, ::std::optional<double> atol, ::std::optional<double> rtol, bool hermitian, at::Tensor & out);
+TORCH_API at::Tensor linalg_pinv(const at::Tensor & self, double rcond, bool hermitian=false);
+TORCH_API at::Tensor & linalg_pinv_out(const at::Tensor & self, double rcond, bool hermitian, at::Tensor & out);
+TORCH_API at::Tensor linalg_pinv(const at::Tensor & self, const at::Tensor & rcond, bool hermitian=false);
+TORCH_API at::Tensor & linalg_pinv_out(const at::Tensor & self, const at::Tensor & rcond, bool hermitian, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_qr.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_qr.h
new file mode 100644
index 0000000000000000000000000000000000000000..60721caa16c157f4f6a02ca6ae043cb3f63015ca
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_qr.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/linalg_qr_ops.h>
+
+namespace at {
+
+
+// aten::linalg_qr(Tensor A, str mode='reduced') -> (Tensor Q, Tensor R)
+inline ::std::tuple<at::Tensor,at::Tensor> linalg_qr(const at::Tensor & A, c10::string_view mode="reduced") {
+    return at::_ops::linalg_qr::call(A, mode);
+}
+
+// aten::linalg_qr.out(Tensor A, str mode='reduced', *, Tensor(a!) Q, Tensor(b!) R) -> (Tensor(a!) Q, Tensor(b!) R)
+inline ::std::tuple<at::Tensor &,at::Tensor &> linalg_qr_out(at::Tensor & Q, at::Tensor & R, const at::Tensor & A, c10::string_view mode="reduced") {
+    return at::_ops::linalg_qr_out::call(A, mode, Q, R);
+}
+// aten::linalg_qr.out(Tensor A, str mode='reduced', *, Tensor(a!) Q, Tensor(b!) R) -> (Tensor(a!) Q, Tensor(b!) R)
+inline ::std::tuple<at::Tensor &,at::Tensor &> linalg_qr_outf(const at::Tensor & A, c10::string_view mode, at::Tensor & Q, at::Tensor & R) {
+    return at::_ops::linalg_qr_out::call(A, mode, Q, R);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve.h
new file mode 100644
index 0000000000000000000000000000000000000000..89e38607cc61b5ba036f14cf0a98ba6e3e8b551f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/linalg_solve_ops.h>
+
+namespace at {
+
+
+// aten::linalg_solve(Tensor A, Tensor B, *, bool left=True) -> Tensor
+inline at::Tensor linalg_solve(const at::Tensor & A, const at::Tensor & B, bool left=true) {
+    return at::_ops::linalg_solve::call(A, B, left);
+}
+
+// aten::linalg_solve.out(Tensor A, Tensor B, *, bool left=True, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_solve_out(at::Tensor & out, const at::Tensor & A, const at::Tensor & B, bool left=true) {
+    return at::_ops::linalg_solve_out::call(A, B, left, out);
+}
+// aten::linalg_solve.out(Tensor A, Tensor B, *, bool left=True, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_solve_outf(const at::Tensor & A, const at::Tensor & B, bool left, at::Tensor & out) {
+    return at::_ops::linalg_solve_out::call(A, B, left, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..55c6a75315bbaa1a5f311bc5ebca127a6463153b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 linalg_solve(const at::Tensor & A, const at::Tensor & B, bool left=true);
+TORCH_API at::Tensor & linalg_solve_out(at::Tensor & out, const at::Tensor & A, const at::Tensor & B, bool left=true);
+TORCH_API at::Tensor & linalg_solve_outf(const at::Tensor & A, const at::Tensor & B, bool left, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_ex_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_ex_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..989ae736b8212764fc1c9b5a9a4bf8a1512e00e4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_ex_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API linalg_solve_ex {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_solve_ex";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "linalg_solve_ex(Tensor A, Tensor B, *, bool left=True, bool check_errors=False) -> (Tensor result, Tensor info)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & A, const at::Tensor & B, bool left, bool check_errors);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & A, const at::Tensor & B, bool left, bool check_errors);
+};
+
+struct TORCH_API linalg_solve_ex_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, const at::Tensor &, bool, bool, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_solve_ex";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "linalg_solve_ex.out(Tensor A, Tensor B, *, bool left=True, bool check_errors=False, Tensor(a!) result, Tensor(b!) info) -> (Tensor(a!) result, Tensor(b!) info)";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & A, const at::Tensor & B, bool left, bool check_errors, at::Tensor & result, at::Tensor & info);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & A, const at::Tensor & B, bool left, bool check_errors, at::Tensor & result, at::Tensor & info);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b835ce4375b7e49ffecac91a03e8495a3fc07785
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_solve(const at::Tensor & A, const at::Tensor & B, bool left=true);
+TORCH_API at::Tensor & linalg_solve_out(const at::Tensor & A, const at::Tensor & B, bool left, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_triangular_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_triangular_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..086e1d96730cd0fc12a9286cc88f8974b950213a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_triangular_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_solve_triangular(const at::Tensor & self, const at::Tensor & B, bool upper, bool left=true, bool unitriangular=false);
+TORCH_API at::Tensor & linalg_solve_triangular_out(const at::Tensor & self, const at::Tensor & B, bool upper, bool left, bool unitriangular, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_triangular_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_triangular_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..5143e6503e7227b0b0e8e6bbd3d02372647a18ad
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linalg_solve_triangular_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API linalg_solve_triangular_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, bool, bool, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_solve_triangular";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "linalg_solve_triangular.out(Tensor self, Tensor B, *, bool upper, bool left=True, bool unitriangular=False, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & B, bool upper, bool left, bool unitriangular, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & B, bool upper, bool left, bool unitriangular, at::Tensor & out);
+};
+
+struct TORCH_API linalg_solve_triangular {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, bool, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::linalg_solve_triangular";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "linalg_solve_triangular(Tensor self, Tensor B, *, bool upper, bool left=True, bool unitriangular=False) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & B, bool upper, bool left, bool unitriangular);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & B, bool upper, bool left, bool unitriangular);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linear_backward_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linear_backward_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8b557b7d1e6de0f26078e4639ad1162397f2284b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/linear_backward_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> linear_backward_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & self, const at::Tensor & grad_output, const at::Tensor & weight, ::std::array<bool,3> output_mask);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> linear_backward_outf(const at::Tensor & self, const at::Tensor & grad_output, const at::Tensor & weight, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log10.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log10.h
new file mode 100644
index 0000000000000000000000000000000000000000..56b192384d601333ce5472343ac6ffa090bcb2ef
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log10.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/log10_ops.h>
+
+namespace at {
+
+
+// aten::log10(Tensor self) -> Tensor
+inline at::Tensor log10(const at::Tensor & self) {
+    return at::_ops::log10::call(self);
+}
+
+// aten::log10_(Tensor(a!) self) -> Tensor(a!)
+inline at::Tensor & log10_(at::Tensor & self) {
+    return at::_ops::log10_::call(self);
+}
+
+// aten::log10.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & log10_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::log10_out::call(self, out);
+}
+// aten::log10.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & log10_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::log10_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log10_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log10_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f68045a0f773d5101c4036a15ea7b0d26e8d0f38
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log10_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 log10(const at::Tensor & self);
+TORCH_API at::Tensor & log10_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log1p_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log1p_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9478df2663bbc6381981135f40cdb65112e888a8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log1p_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 log1p(const at::Tensor & self);
+TORCH_API at::Tensor & log1p_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & log1p_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & log1p_(at::Tensor & self);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log1p_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log1p_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..f3b8ce9c20e90bd2fd7cd33ad1596422f0ab5669
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log1p_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_log1p : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log2_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log2_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8d47653a13a5a536664217e71b1992b2e0c8d041
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log2_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 log2 {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::log2";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "log2(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 log2_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::log2_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "log2_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API log2_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 const char* name = "aten::log2";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "log2.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5d90cb1da6e4c6bba3c7b211270c1b6323e43fe2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 log(const at::Tensor & self);
+TORCH_API at::Tensor & log_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & log_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & log_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..c3413eb82f38fa4ae414a8c69c5ea3e8d27324d0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/log_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_log_out : public at::meta::structured_log {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_sigmoid_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_sigmoid_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3e0b6d71633c3a1a3e0cb18f7b62dba99e4958db
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_sigmoid_backward_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 log_sigmoid_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & buffer);
+TORCH_API at::Tensor & log_sigmoid_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & buffer);
+TORCH_API at::Tensor & log_sigmoid_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & buffer, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_sigmoid_forward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_sigmoid_forward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..1e2a188b58db6756af9402e762b3517656b85f6a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_sigmoid_forward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 log_sigmoid_forward_output {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::log_sigmoid_forward";
+  static constexpr const char* overload_name = "output";
+  static constexpr const char* schema_str = "log_sigmoid_forward.output(Tensor self, *, Tensor(a!) output, Tensor(b!) buffer) -> (Tensor(a!), Tensor(b!))";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, at::Tensor & output, at::Tensor & buffer);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Tensor & output, at::Tensor & buffer);
+};
+
+struct TORCH_API log_sigmoid_forward {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::log_sigmoid_forward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "log_sigmoid_forward(Tensor self) -> (Tensor output, Tensor buffer)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_softmax_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_softmax_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..8fd4932c43a2334381351fbe566c28bfc25ec99c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/log_softmax_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 log_softmax(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & log_softmax_out(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+TORCH_API at::Tensor log_softmax(const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_and_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_and_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..29e594c7dfdb46fb0fcd474187b291ff97588823
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_and_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor logical_and(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & logical_and_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_not_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_not_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a000b488404925617800d6c524af4317cd5605aa
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_not_cuda_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & logical_not_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & logical_not_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_or_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_or_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b4335216cffc4ab8ee9b4129774cb0e58c47523d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_or_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor logical_or(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & logical_or_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_or_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_or_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..4b4f80ee6bee9fb3777589d8bc6c7bfdaa5a138b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_or_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 logical_or(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & logical_or_(at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & logical_or_out(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_xor_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_xor_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..e2ff73f5622426b5794a72ec7aae82c89241b4cb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logical_xor_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 logical_xor {
+  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 const char* name = "aten::logical_xor";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "logical_xor(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 logical_xor_ {
+  using schema = at::Tensor & (at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::logical_xor_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "logical_xor_(Tensor(a!) self, Tensor other) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Tensor & other);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Tensor & other);
+};
+
+struct TORCH_API logical_xor_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 const char* name = "aten::logical_xor";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "logical_xor.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logit_backward_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logit_backward_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..7a559eb214619b8eead2b1e266461bc4f0322f88
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logit_backward_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 logit_backward(const at::Tensor & grad_output, const at::Tensor & self, ::std::optional<double> eps=::std::nullopt);
+TORCH_API at::Tensor & logit_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, ::std::optional<double> eps=::std::nullopt);
+TORCH_API at::Tensor & logit_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, ::std::optional<double> eps, at::Tensor & grad_input);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logsumexp_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logsumexp_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..83e88edaaae6c3504c964f2871811e2d7ba1b7fd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logsumexp_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 logsumexp(const at::Tensor & self, at::IntArrayRef dim, bool keepdim=false);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logsumexp_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logsumexp_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..00bc215c57d3dc9032bb5cc1b05579352ed5ec4e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logsumexp_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 logsumexp(const at::Tensor & self, at::IntArrayRef dim, bool keepdim=false);
+TORCH_API at::Tensor & logsumexp_out(const at::Tensor & self, at::IntArrayRef dim, bool keepdim, at::Tensor & out);
+TORCH_API at::Tensor logsumexp(const at::Tensor & self, at::DimnameList dim, bool keepdim=false);
+TORCH_API at::Tensor & logsumexp_out(const at::Tensor & self, at::DimnameList dim, bool keepdim, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logsumexp_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logsumexp_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..a94ce68456d8def41d64f018153a0c00f47d7056
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/logsumexp_ops.h
@@ -0,0 +1,67 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 logsumexp {
+  using schema = at::Tensor (const at::Tensor &, at::IntArrayRef, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::logsumexp";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "logsumexp(Tensor self, int[1] dim, bool keepdim=False) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::IntArrayRef dim, bool keepdim);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef dim, bool keepdim);
+};
+
+struct TORCH_API logsumexp_out {
+  using schema = at::Tensor & (const at::Tensor &, at::IntArrayRef, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::logsumexp";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "logsumexp.out(Tensor self, int[1] dim, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, at::IntArrayRef dim, bool keepdim, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef dim, bool keepdim, at::Tensor & out);
+};
+
+struct TORCH_API logsumexp_names {
+  using schema = at::Tensor (const at::Tensor &, at::DimnameList, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::logsumexp";
+  static constexpr const char* overload_name = "names";
+  static constexpr const char* schema_str = "logsumexp.names(Tensor self, Dimname[1] dim, bool keepdim=False) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::DimnameList dim, bool keepdim);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::DimnameList dim, bool keepdim);
+};
+
+struct TORCH_API logsumexp_names_out {
+  using schema = at::Tensor & (const at::Tensor &, at::DimnameList, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::logsumexp";
+  static constexpr const char* overload_name = "names_out";
+  static constexpr const char* schema_str = "logsumexp.names_out(Tensor self, Dimname[1] dim, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, at::DimnameList dim, bool keepdim, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::DimnameList dim, bool keepdim, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lstm_cell_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lstm_cell_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..45481d76a8a8e61bb3c261cd778f2f4ced6ea315
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lstm_cell_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 lstm_cell {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, at::TensorList, const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::lstm_cell";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "lstm_cell(Tensor input, Tensor[] hx, Tensor w_ih, Tensor w_hh, Tensor? b_ih=None, Tensor? b_hh=None) -> (Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & input, at::TensorList hx, const at::Tensor & w_ih, const at::Tensor & w_hh, const ::std::optional<at::Tensor> & b_ih, const ::std::optional<at::Tensor> & b_hh);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, at::TensorList hx, const at::Tensor & w_ih, const at::Tensor & w_hh, const ::std::optional<at::Tensor> & b_ih, const ::std::optional<at::Tensor> & b_hh);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lstm_mps_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lstm_mps_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..0d2647c503125151356f86af10810ac88bc3d29b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lstm_mps_backward_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void lstm_mps_backward_out(const ::std::optional<at::Tensor> & grad_y, const ::std::optional<at::Tensor> & grad_hy, const ::std::optional<at::Tensor> & grad_cy, const at::Tensor & z_state, const at::Tensor & cell_state_fwd, const at::Tensor & input, const at::Tensor & layersOutputs, at::TensorList hx, at::TensorList params, bool has_biases, int64_t num_layers, double dropout, bool train, bool bidirectional, bool batch_first, at::Tensor & out0, at::TensorList out1, at::TensorList out2);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lu_solve_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lu_solve_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e15a25c12e713c201608e6f308eeea5644863a95
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lu_solve_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 lu_solve(const at::Tensor & self, const at::Tensor & LU_data, const at::Tensor & LU_pivots);
+TORCH_API at::Tensor & lu_solve_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & LU_data, const at::Tensor & LU_pivots);
+TORCH_API at::Tensor & lu_solve_outf(const at::Tensor & self, const at::Tensor & LU_data, const at::Tensor & LU_pivots, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lu_solve_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lu_solve_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..f36976b9578f00896a72b5805818e544d40c44b0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lu_solve_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 lu_solve_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 const char* name = "aten::lu_solve";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "lu_solve.out(Tensor self, Tensor LU_data, Tensor LU_pivots, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & LU_data, const at::Tensor & LU_pivots, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & LU_data, const at::Tensor & LU_pivots, at::Tensor & out);
+};
+
+struct TORCH_API lu_solve {
+  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 const char* name = "aten::lu_solve";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "lu_solve(Tensor self, Tensor LU_data, Tensor LU_pivots) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & LU_data, const at::Tensor & LU_pivots);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & LU_data, const at::Tensor & LU_pivots);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lu_unpack_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lu_unpack_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..97e45aaec06d06031b8827ef27aff6bf64206c59
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/lu_unpack_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_lu_unpack : public at::impl::MetaBase {
+
+
+    void meta(const at::Tensor & LU_data, const at::Tensor & LU_pivots, bool unpack_data, bool unpack_pivots);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mT_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mT_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..04134c06383ebccc4376c3c9c46a624ed982cdf1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mT_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 mT(const at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/margin_ranking_loss_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/margin_ranking_loss_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..dc70e2b03487e1d507a8299df2d2040186634931
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/margin_ranking_loss_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 margin_ranking_loss {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, double, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::margin_ranking_loss";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "margin_ranking_loss(Tensor input1, Tensor input2, Tensor target, float margin=0.0, int reduction=Mean) -> Tensor";
+  static at::Tensor call(const at::Tensor & input1, const at::Tensor & input2, const at::Tensor & target, double margin, int64_t reduction);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input1, const at::Tensor & input2, const at::Tensor & target, double margin, int64_t reduction);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/masked_fill.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/masked_fill.h
new file mode 100644
index 0000000000000000000000000000000000000000..618408ddde0192accb07329771b93cfe378c0329
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/masked_fill.h
@@ -0,0 +1,59 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/masked_fill_ops.h>
+
+namespace at {
+
+
+// aten::masked_fill.Scalar(Tensor self, Tensor mask, Scalar value) -> Tensor
+inline at::Tensor masked_fill(const at::Tensor & self, const at::Tensor & mask, const at::Scalar & value) {
+    return at::_ops::masked_fill_Scalar::call(self, mask, value);
+}
+
+// aten::masked_fill.Tensor(Tensor self, Tensor mask, Tensor value) -> Tensor
+inline at::Tensor masked_fill(const at::Tensor & self, const at::Tensor & mask, const at::Tensor & value) {
+    return at::_ops::masked_fill_Tensor::call(self, mask, value);
+}
+
+// aten::masked_fill.Scalar_out(Tensor self, Tensor mask, Scalar value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & masked_fill_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mask, const at::Scalar & value) {
+    return at::_ops::masked_fill_Scalar_out::call(self, mask, value, out);
+}
+// aten::masked_fill.Scalar_out(Tensor self, Tensor mask, Scalar value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & masked_fill_outf(const at::Tensor & self, const at::Tensor & mask, const at::Scalar & value, at::Tensor & out) {
+    return at::_ops::masked_fill_Scalar_out::call(self, mask, value, out);
+}
+
+// aten::masked_fill.Tensor_out(Tensor self, Tensor mask, Tensor value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & masked_fill_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mask, const at::Tensor & value) {
+    return at::_ops::masked_fill_Tensor_out::call(self, mask, value, out);
+}
+// aten::masked_fill.Tensor_out(Tensor self, Tensor mask, Tensor value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & masked_fill_outf(const at::Tensor & self, const at::Tensor & mask, const at::Tensor & value, at::Tensor & out) {
+    return at::_ops::masked_fill_Tensor_out::call(self, mask, value, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/masked_fill_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/masked_fill_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9405209a7791e002b23972d1bcd969e0c2b743ca
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/masked_fill_cpu_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & masked_fill_(at::Tensor & self, const at::Tensor & mask, const at::Scalar & value);
+TORCH_API at::Tensor & masked_fill_(at::Tensor & self, const at::Tensor & mask, const at::Tensor & value);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/masked_scatter.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/masked_scatter.h
new file mode 100644
index 0000000000000000000000000000000000000000..33fd4e2a2f266e2b9eaaee7856e1f742f4774d8b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/masked_scatter.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/masked_scatter_ops.h>
+
+namespace at {
+
+
+// aten::masked_scatter(Tensor self, Tensor mask, Tensor source) -> Tensor
+inline at::Tensor masked_scatter(const at::Tensor & self, const at::Tensor & mask, const at::Tensor & source) {
+    return at::_ops::masked_scatter::call(self, mask, source);
+}
+
+// aten::masked_scatter.out(Tensor self, Tensor mask, Tensor source, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & masked_scatter_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mask, const at::Tensor & source) {
+    return at::_ops::masked_scatter_out::call(self, mask, source, out);
+}
+// aten::masked_scatter.out(Tensor self, Tensor mask, Tensor source, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & masked_scatter_outf(const at::Tensor & self, const at::Tensor & mask, const at::Tensor & source, at::Tensor & out) {
+    return at::_ops::masked_scatter_out::call(self, mask, source, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/matmul_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/matmul_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8314e62c897f2b43d759c75326b99a46b52192ad
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/matmul_backward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 matmul_backward {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, ::std::array<bool,2>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::matmul_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "matmul_backward(Tensor grad, Tensor self, Tensor other, bool[2] mask) -> (Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & grad, const at::Tensor & self, const at::Tensor & other, ::std::array<bool,2> mask);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad, const at::Tensor & self, const at::Tensor & other, ::std::array<bool,2> mask);
+};
+
+struct TORCH_API matmul_backward_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, const at::Tensor &, const at::Tensor &, ::std::array<bool,2>, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::matmul_backward";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "matmul_backward.out(Tensor grad, Tensor self, Tensor other, bool[2] mask, *, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & grad, const at::Tensor & self, const at::Tensor & other, ::std::array<bool,2> mask, at::Tensor & out0, at::Tensor & out1);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad, const at::Tensor & self, const at::Tensor & other, ::std::array<bool,2> mask, at::Tensor & out0, at::Tensor & out1);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c1cf97e2590f03c5b30f43b8ba77395f78a364f2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,at::Tensor> max(const at::Tensor & self, at::Dimname dim, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> max_out(at::Tensor & max, at::Tensor & max_values, const at::Tensor & self, at::Dimname dim, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> max_outf(const at::Tensor & self, at::Dimname dim, bool keepdim, at::Tensor & max, at::Tensor & max_values);
+TORCH_API at::Tensor max(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & max_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & max_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_pool1d_with_indices_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_pool1d_with_indices_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..8419b5f0b7ec19f7023c4da9b96e8d76ad4b71c0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_pool1d_with_indices_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> max_pool1d_with_indices(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, at::IntArrayRef dilation=1, bool ceil_mode=false);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_pool2d_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_pool2d_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..8bd20d4be40130f5e056e41e48cfff82bf553fad
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_pool2d_backward_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & max_pool2d_backward_out(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, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_pool3d_with_indices_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_pool3d_with_indices_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8443d43dafc5f1cd3257460e7f9cfa3ce25a5381
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/max_pool3d_with_indices_backward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 max_pool3d_with_indices_backward_grad_input {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, at::IntArrayRef, at::IntArrayRef, at::IntArrayRef, at::IntArrayRef, bool, const at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::max_pool3d_with_indices_backward";
+  static constexpr const char* overload_name = "grad_input";
+  static constexpr const char* schema_str = "max_pool3d_with_indices_backward.grad_input(Tensor grad_output, Tensor self, int[3] kernel_size, int[3] stride, int[3] padding, int[3] dilation, bool ceil_mode, Tensor indices, *, Tensor(a!) grad_input) -> Tensor(a!)";
+  static at::Tensor & call(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);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, 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);
+};
+
+struct TORCH_API max_pool3d_with_indices_backward {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, at::IntArrayRef, at::IntArrayRef, at::IntArrayRef, at::IntArrayRef, bool, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::max_pool3d_with_indices_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "max_pool3d_with_indices_backward(Tensor grad_output, Tensor self, int[3] kernel_size, int[3] stride, int[3] padding, int[3] dilation, bool ceil_mode, Tensor indices) -> Tensor";
+  static at::Tensor call(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);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, 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);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/maximum_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/maximum_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..82d12c03213d68e17f1a77d7cc03d44cf663da73
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/maximum_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/maximum_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_maximum_out : public at::meta::structured_maximum {
+void impl(const at::Tensor & self, const at::Tensor & other, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mean_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mean_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..fb8b34a3a8527e41133c8e6d5293dde599e8a112
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mean_native.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/mean_meta.h>
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor mean(const at::Tensor & self, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & mean_dtype_out(const at::Tensor & self, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+struct TORCH_API structured_mean_out : public at::meta::structured_mean_dim {
+void impl(const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, const at::Tensor & out);
+};
+TORCH_API at::Tensor mean_quantized_cpu(const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim=false, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & mean_out_quantized_cpu(const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+TORCH_API at::Tensor mean(const at::Tensor & self, at::DimnameList dim, bool keepdim=false, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor & mean_out(const at::Tensor & self, at::DimnameList dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mean_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mean_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..ce974cc26b444d9c9531e8582864e304ab6cf44a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mean_ops.h
@@ -0,0 +1,89 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 mean {
+  using schema = at::Tensor (const at::Tensor &, ::std::optional<at::ScalarType>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::mean";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "mean(Tensor self, *, ScalarType? dtype=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, ::std::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<at::ScalarType> dtype);
+};
+
+struct TORCH_API mean_dtype_out {
+  using schema = at::Tensor & (const at::Tensor &, ::std::optional<at::ScalarType>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::mean";
+  static constexpr const char* overload_name = "dtype_out";
+  static constexpr const char* schema_str = "mean.dtype_out(Tensor self, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+};
+
+struct TORCH_API mean_dim {
+  using schema = at::Tensor (const at::Tensor &, at::OptionalIntArrayRef, bool, ::std::optional<at::ScalarType>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::mean";
+  static constexpr const char* overload_name = "dim";
+  static constexpr const char* schema_str = "mean.dim(Tensor self, int[1]? dim, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+};
+
+struct TORCH_API mean_out {
+  using schema = at::Tensor & (const at::Tensor &, at::OptionalIntArrayRef, bool, ::std::optional<at::ScalarType>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::mean";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "mean.out(Tensor self, int[1]? dim, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+};
+
+struct TORCH_API mean_names_dim {
+  using schema = at::Tensor (const at::Tensor &, at::DimnameList, bool, ::std::optional<at::ScalarType>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::mean";
+  static constexpr const char* overload_name = "names_dim";
+  static constexpr const char* schema_str = "mean.names_dim(Tensor self, Dimname[1] dim, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::DimnameList dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::DimnameList dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+};
+
+struct TORCH_API mean_names_out {
+  using schema = at::Tensor & (const at::Tensor &, at::DimnameList, bool, ::std::optional<at::ScalarType>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::mean";
+  static constexpr const char* overload_name = "names_out";
+  static constexpr const char* schema_str = "mean.names_out(Tensor self, Dimname[1] dim, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, at::DimnameList dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::DimnameList dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/median.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/median.h
new file mode 100644
index 0000000000000000000000000000000000000000..d223addd2b27bda52d4b3ac16b5cb80b9d7834b9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/median.h
@@ -0,0 +1,73 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/median_ops.h>
+
+namespace at {
+
+
+// aten::median(Tensor self) -> Tensor
+inline at::Tensor median(const at::Tensor & self) {
+    return at::_ops::median::call(self);
+}
+
+// aten::median.dim(Tensor self, int dim, bool keepdim=False) -> (Tensor values, Tensor indices)
+inline ::std::tuple<at::Tensor,at::Tensor> median(const at::Tensor & self, int64_t dim, bool keepdim=false) {
+    return at::_ops::median_dim::call(self, dim, keepdim);
+}
+
+// aten::median.dim_values(Tensor self, int dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> median_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, int64_t dim, bool keepdim=false) {
+    return at::_ops::median_dim_values::call(self, dim, keepdim, values, indices);
+}
+// aten::median.dim_values(Tensor self, int dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> median_outf(const at::Tensor & self, int64_t dim, bool keepdim, at::Tensor & values, at::Tensor & indices) {
+    return at::_ops::median_dim_values::call(self, dim, keepdim, values, indices);
+}
+
+// aten::median.names_dim(Tensor self, Dimname dim, bool keepdim=False) -> (Tensor values, Tensor indices)
+inline ::std::tuple<at::Tensor,at::Tensor> median(const at::Tensor & self, at::Dimname dim, bool keepdim=false) {
+    return at::_ops::median_names_dim::call(self, dim, keepdim);
+}
+
+// aten::median.names_dim_values(Tensor self, Dimname dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> median_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, at::Dimname dim, bool keepdim=false) {
+    return at::_ops::median_names_dim_values::call(self, dim, keepdim, values, indices);
+}
+// aten::median.names_dim_values(Tensor self, Dimname dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)
+inline ::std::tuple<at::Tensor &,at::Tensor &> median_outf(const at::Tensor & self, at::Dimname dim, bool keepdim, at::Tensor & values, at::Tensor & indices) {
+    return at::_ops::median_names_dim_values::call(self, dim, keepdim, values, indices);
+}
+
+// aten::median.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & median_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::median_out::call(self, out);
+}
+// aten::median.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & median_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::median_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/min_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/min_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6e92fff9f63a0124d34478b37029217f6a06e125
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/min_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> min(const at::Tensor & self, int64_t dim, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> min_out(at::Tensor & min, at::Tensor & min_indices, const at::Tensor & self, int64_t dim, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> min_outf(const at::Tensor & self, int64_t dim, bool keepdim, at::Tensor & min, at::Tensor & min_indices);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/miopen_convolution_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/miopen_convolution_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..85f0e0f71c0eabf5c60e970ed075cfa158c550e9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/miopen_convolution_cuda_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 miopen_convolution(const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, bool benchmark, bool deterministic);
+TORCH_API at::Tensor miopen_convolution_symint(const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, bool benchmark, bool deterministic);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/miopen_ctc_loss_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/miopen_ctc_loss_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5c64d2b2368633c2dbf714695324af76f89c0ed7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/miopen_ctc_loss_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 &> miopen_ctc_loss_out(at::Tensor & out0, at::Tensor & out1, const at::Tensor & log_probs, const at::Tensor & targets, at::IntArrayRef input_lengths, at::IntArrayRef target_lengths, int64_t blank, bool deterministic, bool zero_infinity);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> miopen_ctc_loss_outf(const at::Tensor & log_probs, const at::Tensor & targets, at::IntArrayRef input_lengths, at::IntArrayRef target_lengths, int64_t blank, bool deterministic, bool zero_infinity, at::Tensor & out0, at::Tensor & out1);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/miopen_rnn_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/miopen_rnn_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..71905a19c7a936e8b436f8470cfd851bb32ddb90
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/miopen_rnn_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> miopen_rnn_out(const at::Tensor & input, at::TensorList weight, int64_t weight_stride0, const at::Tensor & hx, const ::std::optional<at::Tensor> & cx, int64_t mode, int64_t hidden_size, int64_t num_layers, bool batch_first, double dropout, bool train, bool bidirectional, at::IntArrayRef batch_sizes, const ::std::optional<at::Tensor> & dropout_state, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3, at::Tensor & out4);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,at::Tensor> miopen_rnn(const at::Tensor & input, at::TensorList weight, int64_t weight_stride0, const at::Tensor & hx, const ::std::optional<at::Tensor> & cx, int64_t mode, int64_t hidden_size, int64_t num_layers, bool batch_first, double dropout, bool train, bool bidirectional, at::IntArrayRef batch_sizes, const ::std::optional<at::Tensor> & dropout_state);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..0f44f489103741f17f6354c08d1b31a591a08ab6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_backward_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 mish_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 const char* name = "aten::mish_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "mish_backward(Tensor grad_output, Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad_output, const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..be11c29134af79cafd808266d908673db8d7ed4a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_mish : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..1fd884449af949c055df01bdb4e458cd34216b14
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 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 meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..c26b77d01d69f07483d02e01ac3312a380d1b4b2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mish_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 mish {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::mish";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "mish(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 mish_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::mish_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "mish_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API mish_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 const char* name = "aten::mish";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "mish.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_convolution_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_convolution_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..d53a2f0c2740b2fbd08fe31e9b5d10583b0082f1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_convolution_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 mkldnn_convolution(const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups);
+TORCH_API at::Tensor & mkldnn_convolution_out_symint(const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_linear_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_linear_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..70703049873b969a065f99357038b97fa20e2833
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_linear_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & mkldnn_linear_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias={});
+TORCH_API at::Tensor & mkldnn_linear_outf(const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_max_pool3d_backward_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_max_pool3d_backward_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..91fa08b5632f8c67e338ae712a30832b221091f6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_max_pool3d_backward_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & 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);
+TORCH_API 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);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_max_pool3d_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_max_pool3d_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..55aeaaafe117e612ee33e7c50b272757208e9db5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_max_pool3d_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & mkldnn_max_pool3d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, at::IntArrayRef dilation=1, bool ceil_mode=false);
+TORCH_API at::Tensor & mkldnn_max_pool3d_outf(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool ceil_mode, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_reorder_conv2d_weight.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_reorder_conv2d_weight.h
new file mode 100644
index 0000000000000000000000000000000000000000..b89a389c74bc14787a0634b51319d2c919b9cc6e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_reorder_conv2d_weight.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/mkldnn_reorder_conv2d_weight_ops.h>
+
+namespace at {
+
+
+// aten::mkldnn_reorder_conv2d_weight(Tensor self, SymInt[2] padding=0, SymInt[2] stride=1, SymInt[2] dilation=1, SymInt groups=1, SymInt[]? input_size=None) -> Tensor
+inline at::Tensor mkldnn_reorder_conv2d_weight(const at::Tensor & self, at::IntArrayRef padding=0, at::IntArrayRef stride=1, at::IntArrayRef dilation=1, int64_t groups=1, at::OptionalIntArrayRef input_size=::std::nullopt) {
+    return at::_ops::mkldnn_reorder_conv2d_weight::call(self, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, input_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*input_size)) : ::std::nullopt);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor mkldnn_reorder_conv2d_weight(const at::Tensor & self, at::IntArrayRef padding=0, at::IntArrayRef stride=1, at::IntArrayRef dilation=1, int64_t groups=1, at::OptionalIntArrayRef input_size=::std::nullopt) {
+    return at::_ops::mkldnn_reorder_conv2d_weight::call(self, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, input_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*input_size)) : ::std::nullopt);
+  }
+}
+
+// aten::mkldnn_reorder_conv2d_weight(Tensor self, SymInt[2] padding=0, SymInt[2] stride=1, SymInt[2] dilation=1, SymInt groups=1, SymInt[]? input_size=None) -> Tensor
+inline at::Tensor mkldnn_reorder_conv2d_weight_symint(const at::Tensor & self, c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef dilation=c10::SymInt(1), c10::SymInt groups=1, at::OptionalSymIntArrayRef input_size=::std::nullopt) {
+    return at::_ops::mkldnn_reorder_conv2d_weight::call(self, padding, stride, dilation, groups, input_size);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor mkldnn_reorder_conv2d_weight(const at::Tensor & self, c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef dilation=c10::SymInt(1), c10::SymInt groups=1, at::OptionalSymIntArrayRef input_size=::std::nullopt) {
+    return at::_ops::mkldnn_reorder_conv2d_weight::call(self, padding, stride, dilation, groups, input_size);
+  }
+}
+
+// aten::mkldnn_reorder_conv2d_weight.out(Tensor self, SymInt[2] padding=0, SymInt[2] stride=1, SymInt[2] dilation=1, SymInt groups=1, SymInt[]? input_size=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & mkldnn_reorder_conv2d_weight_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef padding=0, at::IntArrayRef stride=1, at::IntArrayRef dilation=1, int64_t groups=1, at::OptionalIntArrayRef input_size=::std::nullopt) {
+    return at::_ops::mkldnn_reorder_conv2d_weight_out::call(self, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, input_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*input_size)) : ::std::nullopt, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & mkldnn_reorder_conv2d_weight_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef padding=0, at::IntArrayRef stride=1, at::IntArrayRef dilation=1, int64_t groups=1, at::OptionalIntArrayRef input_size=::std::nullopt) {
+    return at::_ops::mkldnn_reorder_conv2d_weight_out::call(self, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, input_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*input_size)) : ::std::nullopt, out);
+  }
+}
+
+// aten::mkldnn_reorder_conv2d_weight.out(Tensor self, SymInt[2] padding=0, SymInt[2] stride=1, SymInt[2] dilation=1, SymInt groups=1, SymInt[]? input_size=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & mkldnn_reorder_conv2d_weight_outf(const at::Tensor & self, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, at::OptionalIntArrayRef input_size, at::Tensor & out) {
+    return at::_ops::mkldnn_reorder_conv2d_weight_out::call(self, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, input_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*input_size)) : ::std::nullopt, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & mkldnn_reorder_conv2d_weight_outf(const at::Tensor & self, at::IntArrayRef padding, at::IntArrayRef stride, at::IntArrayRef dilation, int64_t groups, at::OptionalIntArrayRef input_size, at::Tensor & out) {
+    return at::_ops::mkldnn_reorder_conv2d_weight_out::call(self, c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(dilation), groups, input_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*input_size)) : ::std::nullopt, out);
+  }
+}
+
+// aten::mkldnn_reorder_conv2d_weight.out(Tensor self, SymInt[2] padding=0, SymInt[2] stride=1, SymInt[2] dilation=1, SymInt groups=1, SymInt[]? input_size=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & mkldnn_reorder_conv2d_weight_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef dilation=c10::SymInt(1), c10::SymInt groups=1, at::OptionalSymIntArrayRef input_size=::std::nullopt) {
+    return at::_ops::mkldnn_reorder_conv2d_weight_out::call(self, padding, stride, dilation, groups, input_size, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & mkldnn_reorder_conv2d_weight_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef dilation=c10::SymInt(1), c10::SymInt groups=1, at::OptionalSymIntArrayRef input_size=::std::nullopt) {
+    return at::_ops::mkldnn_reorder_conv2d_weight_out::call(self, padding, stride, dilation, groups, input_size, out);
+  }
+}
+
+// aten::mkldnn_reorder_conv2d_weight.out(Tensor self, SymInt[2] padding=0, SymInt[2] stride=1, SymInt[2] dilation=1, SymInt groups=1, SymInt[]? input_size=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & mkldnn_reorder_conv2d_weight_symint_outf(const at::Tensor & self, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, at::OptionalSymIntArrayRef input_size, at::Tensor & out) {
+    return at::_ops::mkldnn_reorder_conv2d_weight_out::call(self, padding, stride, dilation, groups, input_size, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & mkldnn_reorder_conv2d_weight_outf(const at::Tensor & self, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, at::OptionalSymIntArrayRef input_size, at::Tensor & out) {
+    return at::_ops::mkldnn_reorder_conv2d_weight_out::call(self, padding, stride, dilation, groups, input_size, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_reorder_conv2d_weight_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_reorder_conv2d_weight_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..bfea7df0f2c3d8004eeaea211782dcc5e1652e54
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mkldnn_reorder_conv2d_weight_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & mkldnn_reorder_conv2d_weight_out_symint(const at::Tensor & self, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, at::OptionalSymIntArrayRef input_size, at::Tensor & out);
+TORCH_API at::Tensor mkldnn_reorder_conv2d_weight(const at::Tensor & self, at::IntArrayRef padding=0, at::IntArrayRef stride=1, at::IntArrayRef dilation=1, int64_t groups=1, at::OptionalIntArrayRef input_size=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mm_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mm_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..aa484a8c2807542af1d168ed3b129440cfb9f722
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mm_cuda_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor mm(const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor & mm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor & mm_outf(const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out);
+TORCH_API at::Tensor mm(const at::Tensor & self, const at::Tensor & mat2, at::ScalarType out_dtype);
+TORCH_API at::Tensor & mm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat2, at::ScalarType out_dtype);
+TORCH_API at::Tensor & mm_outf(const at::Tensor & self, const at::Tensor & mat2, at::ScalarType out_dtype, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mode_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mode_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..74169f11d519a492c2479553524d6f021415a3df
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mode_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> mode_out(const at::Tensor & self, int64_t dim, bool keepdim, at::Tensor & values, at::Tensor & indices);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> mode(const at::Tensor & self, int64_t dim=-1, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> mode(const at::Tensor & self, at::Dimname dim, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> mode_out(const at::Tensor & self, at::Dimname dim, bool keepdim, at::Tensor & values, at::Tensor & indices);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/moveaxis_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/moveaxis_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..39d5366ad61f8df25b505890d1added2d230a7cc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/moveaxis_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 moveaxis(const at::Tensor & self, at::IntArrayRef source, at::IntArrayRef destination);
+TORCH_API at::Tensor moveaxis(const at::Tensor & self, int64_t source, int64_t destination);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/moveaxis_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/moveaxis_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..ea51c3df31edf9fda82a013c2242a6dc185ebc4d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/moveaxis_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 moveaxis(const at::Tensor & self, at::IntArrayRef source, at::IntArrayRef destination);
+TORCH_API at::Tensor moveaxis(const at::Tensor & self, int64_t source, int64_t destination);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/movedim_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/movedim_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..3c8e01ba87c61d58bc26cdb5d54571965bdd36f5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/movedim_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 movedim_intlist {
+  using schema = at::Tensor (const at::Tensor &, at::IntArrayRef, at::IntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::movedim";
+  static constexpr const char* overload_name = "intlist";
+  static constexpr const char* schema_str = "movedim.intlist(Tensor(a) self, int[] source, int[] destination) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, at::IntArrayRef source, at::IntArrayRef destination);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef source, at::IntArrayRef destination);
+};
+
+struct TORCH_API movedim_int {
+  using schema = at::Tensor (const at::Tensor &, int64_t, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::movedim";
+  static constexpr const char* overload_name = "int";
+  static constexpr const char* schema_str = "movedim.int(Tensor(a) self, int source, int destination) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, int64_t source, int64_t destination);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t source, int64_t destination);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mps_convolution_backward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mps_convolution_backward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..d80d3117cfdaf23eac8c9a3c71303b3904914545
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mps_convolution_backward_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> mps_convolution_backward_out_symint(const at::Tensor & self, const at::Tensor & grad_output, const at::Tensor & weight, c10::SymIntArrayRef padding, c10::SymIntArrayRef stride, c10::SymIntArrayRef dilation, c10::SymInt groups, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss.h
new file mode 100644
index 0000000000000000000000000000000000000000..6b8a5ccdaca5f653aafc446384d4b670c9aa1249
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/mse_loss_ops.h>
+
+namespace at {
+
+
+// aten::mse_loss.out(Tensor self, Tensor target, int reduction=Mean, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & mse_loss_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & target, int64_t reduction=at::Reduction::Mean) {
+    return at::_ops::mse_loss_out::call(self, target, reduction, out);
+}
+// aten::mse_loss.out(Tensor self, Tensor target, int reduction=Mean, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & mse_loss_outf(const at::Tensor & self, const at::Tensor & target, int64_t reduction, at::Tensor & out) {
+    return at::_ops::mse_loss_out::call(self, target, reduction, out);
+}
+
+// aten::mse_loss(Tensor self, Tensor target, int reduction=Mean) -> Tensor
+inline at::Tensor mse_loss(const at::Tensor & self, const at::Tensor & target, int64_t reduction=at::Reduction::Mean) {
+    return at::_ops::mse_loss::call(self, target, reduction);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4584bebc8d8ebf26a13517f8a623badb42bf9ae0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss_backward_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 mse_loss_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & target, int64_t reduction);
+TORCH_API at::Tensor & mse_loss_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & target, int64_t reduction);
+TORCH_API at::Tensor & mse_loss_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & target, int64_t reduction, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..d1ac11224ac580d337b8e356e2fdf8ca22390cb6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_mse_loss : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & self, const at::Tensor & target, int64_t reduction);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..c21f17eaee5c12646779c857df5137098ec8b6dc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mse_loss_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/mse_loss_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_mse_loss_out : public at::meta::structured_mse_loss {
+void impl(const at::Tensor & self, const at::Tensor & target, int64_t reduction, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mul_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mul_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..d82c5a403c066e504b0ce51ceba223ef58092374
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mul_native.h
@@ -0,0 +1,48 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/mul_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_mul_out : public at::meta::structured_mul_Tensor {
+void impl(const at::Tensor & self, const at::Tensor & other, const at::Tensor & out);
+};
+TORCH_API at::Tensor NestedTensor_mul_Tensor(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & NestedTensor_mul__Tensor(at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor mul_sparse(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & mul_out_sparse_cpu(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & mul_sparse_(at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & mul_out_sparse_cuda(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor mul_sparse_csr(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & mul_out_sparse_csr(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & mul_sparse_csr_(at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor mkldnn_mul(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & mkldnn_mul_out(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & mkldnn_mul_(at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor mul_zerotensor(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor mul(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & mul_Scalar_out(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & mul_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor NestedTensor_mul_Scalar(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & NestedTensor_mul__Scalar(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor mul_scalar_sparse_csr(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & mul__scalar_sparse_csr(at::Tensor & self, const at::Scalar & other);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multi_margin_loss_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multi_margin_loss_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..869b70582b327e1fb25ce2fda473148b1e79ddc3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multi_margin_loss_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 multi_margin_loss(const at::Tensor & self, const at::Tensor & target, const at::Scalar & p=1, const at::Scalar & margin=1, const ::std::optional<at::Tensor> & weight={}, int64_t reduction=at::Reduction::Mean);
+TORCH_API at::Tensor & multi_margin_loss_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & target, const at::Scalar & p=1, const at::Scalar & margin=1, const ::std::optional<at::Tensor> & weight={}, int64_t reduction=at::Reduction::Mean);
+TORCH_API at::Tensor & multi_margin_loss_outf(const at::Tensor & self, const at::Tensor & target, const at::Scalar & p, const at::Scalar & margin, const ::std::optional<at::Tensor> & weight, int64_t reduction, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multilabel_margin_loss_forward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multilabel_margin_loss_forward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c05f570d1b1559a92b5d752727eb4edf995515ca
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multilabel_margin_loss_forward_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> multilabel_margin_loss_forward(const at::Tensor & self, const at::Tensor & target, int64_t reduction);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> multilabel_margin_loss_forward_out(at::Tensor & output, at::Tensor & is_target, const at::Tensor & self, const at::Tensor & target, int64_t reduction);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> multilabel_margin_loss_forward_outf(const at::Tensor & self, const at::Tensor & target, int64_t reduction, at::Tensor & output, at::Tensor & is_target);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multilabel_margin_loss_forward_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multilabel_margin_loss_forward_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..4ff8510c59f20a52fb6f9629619dc41efe26d184
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multilabel_margin_loss_forward_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> multilabel_margin_loss_forward_cpu(const at::Tensor & self, const at::Tensor & target, int64_t reduction);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> multilabel_margin_loss_forward_out_cpu(const at::Tensor & self, const at::Tensor & target, int64_t reduction, at::Tensor & output, at::Tensor & is_target);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> multilabel_margin_loss_forward_cuda(const at::Tensor & self, const at::Tensor & target, int64_t reduction);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> multilabel_margin_loss_forward_out_cuda(const at::Tensor & self, const at::Tensor & target, int64_t reduction, at::Tensor & output, at::Tensor & is_target);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multinomial.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multinomial.h
new file mode 100644
index 0000000000000000000000000000000000000000..284e3847e820507edd42dc47a0e46458c6508607
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multinomial.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/multinomial_ops.h>
+
+namespace at {
+
+
+// aten::multinomial.out(Tensor self, SymInt num_samples, bool replacement=False, *, Generator? generator=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & multinomial_out(at::Tensor & out, const at::Tensor & self, int64_t num_samples, bool replacement=false, ::std::optional<at::Generator> generator=::std::nullopt) {
+    return at::_ops::multinomial_out::call(self, num_samples, replacement, generator, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & multinomial_out(at::Tensor & out, const at::Tensor & self, int64_t num_samples, bool replacement=false, ::std::optional<at::Generator> generator=::std::nullopt) {
+    return at::_ops::multinomial_out::call(self, num_samples, replacement, generator, out);
+  }
+}
+
+// aten::multinomial.out(Tensor self, SymInt num_samples, bool replacement=False, *, Generator? generator=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & multinomial_outf(const at::Tensor & self, int64_t num_samples, bool replacement, ::std::optional<at::Generator> generator, at::Tensor & out) {
+    return at::_ops::multinomial_out::call(self, num_samples, replacement, generator, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & multinomial_outf(const at::Tensor & self, int64_t num_samples, bool replacement, ::std::optional<at::Generator> generator, at::Tensor & out) {
+    return at::_ops::multinomial_out::call(self, num_samples, replacement, generator, out);
+  }
+}
+
+// aten::multinomial.out(Tensor self, SymInt num_samples, bool replacement=False, *, Generator? generator=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & multinomial_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymInt num_samples, bool replacement=false, ::std::optional<at::Generator> generator=::std::nullopt) {
+    return at::_ops::multinomial_out::call(self, num_samples, replacement, generator, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & multinomial_out(at::Tensor & out, const at::Tensor & self, c10::SymInt num_samples, bool replacement=false, ::std::optional<at::Generator> generator=::std::nullopt) {
+    return at::_ops::multinomial_out::call(self, num_samples, replacement, generator, out);
+  }
+}
+
+// aten::multinomial.out(Tensor self, SymInt num_samples, bool replacement=False, *, Generator? generator=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & multinomial_symint_outf(const at::Tensor & self, c10::SymInt num_samples, bool replacement, ::std::optional<at::Generator> generator, at::Tensor & out) {
+    return at::_ops::multinomial_out::call(self, num_samples, replacement, generator, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & multinomial_outf(const at::Tensor & self, c10::SymInt num_samples, bool replacement, ::std::optional<at::Generator> generator, at::Tensor & out) {
+    return at::_ops::multinomial_out::call(self, num_samples, replacement, generator, out);
+  }
+}
+
+// aten::multinomial(Tensor self, SymInt num_samples, bool replacement=False, *, Generator? generator=None) -> Tensor
+inline at::Tensor multinomial(const at::Tensor & self, int64_t num_samples, bool replacement=false, ::std::optional<at::Generator> generator=::std::nullopt) {
+    return at::_ops::multinomial::call(self, num_samples, replacement, generator);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor multinomial(const at::Tensor & self, int64_t num_samples, bool replacement=false, ::std::optional<at::Generator> generator=::std::nullopt) {
+    return at::_ops::multinomial::call(self, num_samples, replacement, generator);
+  }
+}
+
+// aten::multinomial(Tensor self, SymInt num_samples, bool replacement=False, *, Generator? generator=None) -> Tensor
+inline at::Tensor multinomial_symint(const at::Tensor & self, c10::SymInt num_samples, bool replacement=false, ::std::optional<at::Generator> generator=::std::nullopt) {
+    return at::_ops::multinomial::call(self, num_samples, replacement, generator);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor multinomial(const at::Tensor & self, c10::SymInt num_samples, bool replacement=false, ::std::optional<at::Generator> generator=::std::nullopt) {
+    return at::_ops::multinomial::call(self, num_samples, replacement, generator);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multinomial_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multinomial_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..c178a8bdecdc17df37ffb07176a2cc0b3084a312
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/multinomial_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 multinomial_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymInt, bool, ::std::optional<at::Generator>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::multinomial";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "multinomial.out(Tensor self, SymInt num_samples, bool replacement=False, *, Generator? generator=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymInt num_samples, bool replacement, ::std::optional<at::Generator> generator, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt num_samples, bool replacement, ::std::optional<at::Generator> generator, at::Tensor & out);
+};
+
+struct TORCH_API multinomial {
+  using schema = at::Tensor (const at::Tensor &, c10::SymInt, bool, ::std::optional<at::Generator>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::multinomial";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "multinomial(Tensor self, SymInt num_samples, bool replacement=False, *, Generator? generator=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymInt num_samples, bool replacement, ::std::optional<at::Generator> generator);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt num_samples, bool replacement, ::std::optional<at::Generator> generator);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mv.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mv.h
new file mode 100644
index 0000000000000000000000000000000000000000..a900dfe28762ea310019ddab024ce0e3537e032f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/mv.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/mv_ops.h>
+
+namespace at {
+
+
+// aten::mv(Tensor self, Tensor vec) -> Tensor
+inline at::Tensor mv(const at::Tensor & self, const at::Tensor & vec) {
+    return at::_ops::mv::call(self, vec);
+}
+
+// aten::mv.out(Tensor self, Tensor vec, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & mv_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & vec) {
+    return at::_ops::mv_out::call(self, vec, out);
+}
+// aten::mv.out(Tensor self, Tensor vec, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & mv_outf(const at::Tensor & self, const at::Tensor & vec, at::Tensor & out) {
+    return at::_ops::mv_out::call(self, vec, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/narrow_copy_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/narrow_copy_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..15016503419733bcdab287895a6eaa0ab08d9725
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/narrow_copy_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 narrow_copy(const at::Tensor & self, int64_t dim, int64_t start, int64_t length);
+TORCH_API at::Tensor narrow_copy_symint(const at::Tensor & self, int64_t dim, c10::SymInt start, c10::SymInt length);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_batch_norm_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_batch_norm_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d8e6d25a0280a7a0f7c6e11c5f1459f017f4e579
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_batch_norm_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_batch_norm_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_batch_norm_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..3ef8d14993195d62343e81b86f01d5a70447012c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_batch_norm_native.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> batch_norm_cpu(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 &> batch_norm_cpu_out(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);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> batch_norm_cuda(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 &> batch_norm_cuda_out(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);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> mkldnn_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);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_channel_shuffle_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_channel_shuffle_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..272c077601f5ce9b50ebdb398ae76bab6f1d9077
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_channel_shuffle_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 native_channel_shuffle(const at::Tensor & self, int64_t groups);
+TORCH_API at::Tensor native_channel_shuffle_symint(const at::Tensor & self, c10::SymInt groups);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_channel_shuffle_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_channel_shuffle_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..e19b4dfe215eced99f3e0f0c3b177808d997ccd3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_channel_shuffle_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 math_channel_shuffle(const at::Tensor & self, int64_t groups);
+TORCH_API at::Tensor channel_shuffle_cpu(const at::Tensor & self, int64_t groups);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_group_norm_backward.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_group_norm_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..1f244e78d1786925630595316dbb29bcb84ac854
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_group_norm_backward.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/native_group_norm_backward_ops.h>
+
+namespace at {
+
+
+// aten::native_group_norm_backward(Tensor grad_out, Tensor input, Tensor mean, Tensor rstd, Tensor? weight, SymInt N, SymInt C, SymInt HxW, int group, bool[3] output_mask) -> (Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor> native_group_norm_backward(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, int64_t N, int64_t C, int64_t HxW, int64_t group, ::std::array<bool,3> output_mask) {
+    return at::_ops::native_group_norm_backward::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor,at::Tensor,at::Tensor> native_group_norm_backward(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, int64_t N, int64_t C, int64_t HxW, int64_t group, ::std::array<bool,3> output_mask) {
+    return at::_ops::native_group_norm_backward::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask);
+  }
+}
+
+// aten::native_group_norm_backward(Tensor grad_out, Tensor input, Tensor mean, Tensor rstd, Tensor? weight, SymInt N, SymInt C, SymInt HxW, int group, bool[3] output_mask) -> (Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor> native_group_norm_backward_symint(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, c10::SymInt N, c10::SymInt C, c10::SymInt HxW, int64_t group, ::std::array<bool,3> output_mask) {
+    return at::_ops::native_group_norm_backward::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor,at::Tensor,at::Tensor> native_group_norm_backward(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, c10::SymInt N, c10::SymInt C, c10::SymInt HxW, int64_t group, ::std::array<bool,3> output_mask) {
+    return at::_ops::native_group_norm_backward::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask);
+  }
+}
+
+// aten::native_group_norm_backward.out(Tensor grad_out, Tensor input, Tensor mean, Tensor rstd, Tensor? weight, SymInt N, SymInt C, SymInt HxW, int group, bool[3] output_mask, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_group_norm_backward_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, int64_t N, int64_t C, int64_t HxW, int64_t group, ::std::array<bool,3> output_mask) {
+    return at::_ops::native_group_norm_backward_out::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask, out0, out1, out2);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_group_norm_backward_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, int64_t N, int64_t C, int64_t HxW, int64_t group, ::std::array<bool,3> output_mask) {
+    return at::_ops::native_group_norm_backward_out::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask, out0, out1, out2);
+  }
+}
+
+// aten::native_group_norm_backward.out(Tensor grad_out, Tensor input, Tensor mean, Tensor rstd, Tensor? weight, SymInt N, SymInt C, SymInt HxW, int group, bool[3] output_mask, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_group_norm_backward_outf(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, int64_t N, int64_t C, int64_t HxW, int64_t group, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2) {
+    return at::_ops::native_group_norm_backward_out::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask, out0, out1, out2);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_group_norm_backward_outf(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, int64_t N, int64_t C, int64_t HxW, int64_t group, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2) {
+    return at::_ops::native_group_norm_backward_out::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask, out0, out1, out2);
+  }
+}
+
+// aten::native_group_norm_backward.out(Tensor grad_out, Tensor input, Tensor mean, Tensor rstd, Tensor? weight, SymInt N, SymInt C, SymInt HxW, int group, bool[3] output_mask, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_group_norm_backward_symint_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, c10::SymInt N, c10::SymInt C, c10::SymInt HxW, int64_t group, ::std::array<bool,3> output_mask) {
+    return at::_ops::native_group_norm_backward_out::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask, out0, out1, out2);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_group_norm_backward_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, c10::SymInt N, c10::SymInt C, c10::SymInt HxW, int64_t group, ::std::array<bool,3> output_mask) {
+    return at::_ops::native_group_norm_backward_out::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask, out0, out1, out2);
+  }
+}
+
+// aten::native_group_norm_backward.out(Tensor grad_out, Tensor input, Tensor mean, Tensor rstd, Tensor? weight, SymInt N, SymInt C, SymInt HxW, int group, bool[3] output_mask, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_group_norm_backward_symint_outf(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, c10::SymInt N, c10::SymInt C, c10::SymInt HxW, int64_t group, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2) {
+    return at::_ops::native_group_norm_backward_out::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask, out0, out1, out2);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_group_norm_backward_outf(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, c10::SymInt N, c10::SymInt C, c10::SymInt HxW, int64_t group, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2) {
+    return at::_ops::native_group_norm_backward_out::call(grad_out, input, mean, rstd, weight, N, C, HxW, group, output_mask, out0, out1, out2);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_group_norm_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_group_norm_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8283981af8cc37e7ae0c4b078b466592cfbe26a1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_group_norm_backward_cpu_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> native_group_norm_backward(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, int64_t N, int64_t C, int64_t HxW, int64_t group, ::std::array<bool,3> output_mask);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> native_group_norm_backward_symint(const at::Tensor & grad_out, const at::Tensor & input, const at::Tensor & mean, const at::Tensor & rstd, const ::std::optional<at::Tensor> & weight, c10::SymInt N, c10::SymInt C, c10::SymInt HxW, int64_t group, ::std::array<bool,3> output_mask);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_layer_norm.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_layer_norm.h
new file mode 100644
index 0000000000000000000000000000000000000000..77cc622a7fbd71e6bb4789e0bea08f72e02bac99
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_layer_norm.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/native_layer_norm_ops.h>
+
+namespace at {
+
+
+// aten::native_layer_norm(Tensor input, SymInt[] normalized_shape, Tensor? weight, Tensor? bias, float eps) -> (Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor> native_layer_norm(const at::Tensor & input, at::IntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps) {
+    return at::_ops::native_layer_norm::call(input, c10::fromIntArrayRefSlow(normalized_shape), weight, bias, eps);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor,at::Tensor,at::Tensor> native_layer_norm(const at::Tensor & input, at::IntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps) {
+    return at::_ops::native_layer_norm::call(input, c10::fromIntArrayRefSlow(normalized_shape), weight, bias, eps);
+  }
+}
+
+// aten::native_layer_norm(Tensor input, SymInt[] normalized_shape, Tensor? weight, Tensor? bias, float eps) -> (Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor> native_layer_norm_symint(const at::Tensor & input, c10::SymIntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps) {
+    return at::_ops::native_layer_norm::call(input, normalized_shape, weight, bias, eps);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor,at::Tensor,at::Tensor> native_layer_norm(const at::Tensor & input, c10::SymIntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps) {
+    return at::_ops::native_layer_norm::call(input, normalized_shape, weight, bias, eps);
+  }
+}
+
+// aten::native_layer_norm.out(Tensor input, SymInt[] normalized_shape, Tensor? weight, Tensor? bias, float eps, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_layer_norm_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & input, at::IntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps) {
+    return at::_ops::native_layer_norm_out::call(input, c10::fromIntArrayRefSlow(normalized_shape), weight, bias, eps, out0, out1, out2);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_layer_norm_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & input, at::IntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps) {
+    return at::_ops::native_layer_norm_out::call(input, c10::fromIntArrayRefSlow(normalized_shape), weight, bias, eps, out0, out1, out2);
+  }
+}
+
+// aten::native_layer_norm.out(Tensor input, SymInt[] normalized_shape, Tensor? weight, Tensor? bias, float eps, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_layer_norm_outf(const at::Tensor & input, at::IntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2) {
+    return at::_ops::native_layer_norm_out::call(input, c10::fromIntArrayRefSlow(normalized_shape), weight, bias, eps, out0, out1, out2);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_layer_norm_outf(const at::Tensor & input, at::IntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2) {
+    return at::_ops::native_layer_norm_out::call(input, c10::fromIntArrayRefSlow(normalized_shape), weight, bias, eps, out0, out1, out2);
+  }
+}
+
+// aten::native_layer_norm.out(Tensor input, SymInt[] normalized_shape, Tensor? weight, Tensor? bias, float eps, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_layer_norm_symint_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & input, c10::SymIntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps) {
+    return at::_ops::native_layer_norm_out::call(input, normalized_shape, weight, bias, eps, out0, out1, out2);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_layer_norm_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & input, c10::SymIntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps) {
+    return at::_ops::native_layer_norm_out::call(input, normalized_shape, weight, bias, eps, out0, out1, out2);
+  }
+}
+
+// aten::native_layer_norm.out(Tensor input, SymInt[] normalized_shape, Tensor? weight, Tensor? bias, float eps, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_layer_norm_symint_outf(const at::Tensor & input, c10::SymIntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2) {
+    return at::_ops::native_layer_norm_out::call(input, normalized_shape, weight, bias, eps, out0, out1, out2);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> native_layer_norm_outf(const at::Tensor & input, c10::SymIntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, double eps, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2) {
+    return at::_ops::native_layer_norm_out::call(input, normalized_shape, weight, bias, eps, out0, out1, out2);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_norm_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_norm_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6bca8282d18019c73be40cf5ff0fcb4ad1d95fb9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/native_norm_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & native_norm_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & p=2);
+TORCH_API at::Tensor & native_norm_outf(const at::Tensor & self, const at::Scalar & p, at::Tensor & out);
+TORCH_API at::Tensor & native_norm_out(at::Tensor & out, const at::Tensor & self, const ::std::optional<at::Scalar> & p, at::IntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+TORCH_API at::Tensor & native_norm_outf(const at::Tensor & self, const ::std::optional<at::Scalar> & p, at::IntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/neg.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/neg.h
new file mode 100644
index 0000000000000000000000000000000000000000..e927f225661e893bf437a5075e5d280056bdc7db
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/neg.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/neg_ops.h>
+
+namespace at {
+
+
+// aten::neg(Tensor self) -> Tensor
+inline at::Tensor neg(const at::Tensor & self) {
+    return at::_ops::neg::call(self);
+}
+
+// aten::neg_(Tensor(a!) self) -> Tensor(a!)
+inline at::Tensor & neg_(at::Tensor & self) {
+    return at::_ops::neg_::call(self);
+}
+
+// aten::neg.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & neg_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::neg_out::call(self, out);
+}
+// aten::neg.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & neg_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::neg_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/neg_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/neg_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9cd2605a26581b8d38ecc9bfb714681eae716a7b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/neg_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 neg(const at::Tensor & self);
+TORCH_API at::Tensor & neg_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/negative_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/negative_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..cbe1fc736965887016c4f05cc6bbd25d35394f72
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/negative_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 negative {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::negative";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "negative(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 negative_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::negative_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "negative_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API negative_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 const char* name = "aten::negative";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "negative.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nested_to_padded_tensor_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nested_to_padded_tensor_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..6a9590b77ca2fbbc2e52e66c931106ebba61b8d1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nested_to_padded_tensor_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 nested_to_padded_tensor(const at::Tensor & self, double padding, at::OptionalIntArrayRef output_size=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_empty_strided_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_empty_strided_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..24d96ebe44389925d9ef6cb159b9b15bf20f9d28
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_empty_strided_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & new_empty_strided_out_symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, at::Tensor & out);
+TORCH_API at::Tensor new_empty_strided_symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_full_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_full_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f36e4140d33ea64dbec0717d57897d31ade3b4b1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_full_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,35 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 new_full(const at::Tensor & self, at::IntArrayRef size, const at::Scalar & fill_value, at::TensorOptions options={});
+TORCH_API at::Tensor new_full(const at::Tensor & self, at::IntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+TORCH_API at::Tensor new_full_symint(const at::Tensor & self, c10::SymIntArrayRef size, const at::Scalar & fill_value, at::TensorOptions options={});
+TORCH_API at::Tensor new_full_symint(const at::Tensor & self, c10::SymIntArrayRef size, const at::Scalar & fill_value, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+TORCH_API at::Tensor & new_full_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef size, const at::Scalar & fill_value);
+TORCH_API at::Tensor & new_full_outf(const at::Tensor & self, at::IntArrayRef size, const at::Scalar & fill_value, at::Tensor & out);
+TORCH_API at::Tensor & new_full_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef size, const at::Scalar & fill_value);
+TORCH_API at::Tensor & new_full_symint_outf(const at::Tensor & self, c10::SymIntArrayRef size, const at::Scalar & fill_value, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_ones_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_ones_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..75d558f746184d38727dfd7d9b9305adcfdd722a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_ones_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API new_ones {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::new_ones";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "new_ones(Tensor self, SymInt[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+struct TORCH_API new_ones_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymIntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::new_ones";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "new_ones.out(Tensor self, SymInt[] size, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymIntArrayRef size, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef size, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_zeros_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_zeros_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..a0511e30ac428695a90251e72fb0e7ae5f788663
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/new_zeros_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 new_zeros(const at::Tensor & self, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor & new_zeros_out_symint(const at::Tensor & self, c10::SymIntArrayRef size, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss2d_forward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss2d_forward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e1e51e1ef9f61e0723f5dfe6e1aa310c909048df
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss2d_forward_cpu_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> nll_loss2d_forward(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, int64_t ignore_index);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> nll_loss2d_forward_symint(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss2d_forward_out(at::Tensor & output, at::Tensor & total_weight, const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, int64_t ignore_index);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss2d_forward_outf(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, int64_t ignore_index, at::Tensor & output, at::Tensor & total_weight);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss2d_forward_symint_out(at::Tensor & output, at::Tensor & total_weight, const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss2d_forward_symint_outf(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index, at::Tensor & output, at::Tensor & total_weight);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss2d_forward_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss2d_forward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..44ba2f33d73b49e7c09b25a579a976f12f6a7a40
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss2d_forward_cuda_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> nll_loss2d_forward(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, int64_t ignore_index);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> nll_loss2d_forward_symint(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss2d_forward_out(at::Tensor & output, at::Tensor & total_weight, const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, int64_t ignore_index);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss2d_forward_outf(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, int64_t ignore_index, at::Tensor & output, at::Tensor & total_weight);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss2d_forward_symint_out(at::Tensor & output, at::Tensor & total_weight, const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss2d_forward_symint_outf(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index, at::Tensor & output, at::Tensor & total_weight);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss_forward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss_forward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c824a1161da4eabafc44a4d826abc0b9f6b1f1ea
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss_forward_cpu_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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> nll_loss_forward(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, int64_t ignore_index);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> nll_loss_forward_symint(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss_forward_out(at::Tensor & output, at::Tensor & total_weight, const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, int64_t ignore_index);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss_forward_outf(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, int64_t ignore_index, at::Tensor & output, at::Tensor & total_weight);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss_forward_symint_out(at::Tensor & output, at::Tensor & total_weight, const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> nll_loss_forward_symint_outf(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index, at::Tensor & output, at::Tensor & total_weight);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss_nd_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss_nd_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..98f881b75592872ed8153cf3964edfc729f473c5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nll_loss_nd_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 nll_loss_nd {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, int64_t, c10::SymInt);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::nll_loss_nd";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "nll_loss_nd(Tensor self, Tensor target, Tensor? weight=None, int reduction=Mean, SymInt ignore_index=-100) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & target, const ::std::optional<at::Tensor> & weight, int64_t reduction, c10::SymInt ignore_index);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nonzero_numpy_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nonzero_numpy_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b1b992e6123c87687a028a7ce3e0f12610f0ef93
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nonzero_numpy_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::vector<at::Tensor> nonzero_numpy(const at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nonzero_numpy_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nonzero_numpy_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..daa4908b74937beff0ac223d195230b6a5b6d282
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/nonzero_numpy_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 nonzero_numpy {
+  using schema = ::std::vector<at::Tensor> (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::nonzero_numpy";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "nonzero_numpy(Tensor self) -> Tensor[]";
+  static ::std::vector<at::Tensor> call(const at::Tensor & self);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/norm_except_dim_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/norm_except_dim_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..374dbf7e7dea2fe450e109419239c9ef63e1b625
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/norm_except_dim_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 norm_except_dim {
+  using schema = at::Tensor (const at::Tensor &, int64_t, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::norm_except_dim";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "norm_except_dim(Tensor v, int pow=2, int dim=0) -> Tensor";
+  static at::Tensor call(const at::Tensor & v, int64_t pow, int64_t dim);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & v, int64_t pow, int64_t dim);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/not_equal_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/not_equal_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..2f8c33eeaed155c358518ba3f4f6ddea4038e255
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/not_equal_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,35 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 not_equal(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & not_equal_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & not_equal_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & not_equal_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor not_equal(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & not_equal_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & not_equal_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & not_equal_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/not_equal_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/not_equal_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..472b56b90e6bf46d38a05d519a4f9c7bca09b6ee
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/not_equal_ops.h
@@ -0,0 +1,89 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 not_equal_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 const char* name = "aten::not_equal";
+  static constexpr const char* overload_name = "Scalar_out";
+  static constexpr const char* schema_str = "not_equal.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+};
+
+struct TORCH_API not_equal_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 const char* name = "aten::not_equal";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "not_equal.Scalar(Tensor self, Scalar other) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Scalar & other);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & other);
+};
+
+struct TORCH_API not_equal_Tensor_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 const char* name = "aten::not_equal";
+  static constexpr const char* overload_name = "Tensor_out";
+  static constexpr const char* schema_str = "not_equal.Tensor_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 not_equal_Tensor {
+  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 const char* name = "aten::not_equal";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "not_equal.Tensor(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 not_equal__Scalar {
+  using schema = at::Tensor & (at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::not_equal_";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "not_equal_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Scalar & other);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Scalar & other);
+};
+
+struct TORCH_API not_equal__Tensor {
+  using schema = at::Tensor & (at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::not_equal_";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "not_equal_.Tensor(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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ones_like.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ones_like.h
new file mode 100644
index 0000000000000000000000000000000000000000..09363d5b8526eec591fc5997c31489267876dd20
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ones_like.h
@@ -0,0 +1,49 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/ones_like_ops.h>
+
+namespace at {
+
+
+// aten::ones_like(Tensor self, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor
+inline at::Tensor ones_like(const at::Tensor & self, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt) {
+    return at::_ops::ones_like::call(self, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt(), c10::impl::check_tensor_options_and_extract_memory_format(options, memory_format));
+}
+// aten::ones_like(Tensor self, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor
+inline at::Tensor ones_like(const at::Tensor & self, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format) {
+    return at::_ops::ones_like::call(self, dtype, layout, device, pin_memory, memory_format);
+}
+
+// aten::ones_like.out(Tensor self, *, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & ones_like_out(at::Tensor & out, const at::Tensor & self, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt) {
+    return at::_ops::ones_like_out::call(self, memory_format, out);
+}
+// aten::ones_like.out(Tensor self, *, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & ones_like_outf(const at::Tensor & self, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out) {
+    return at::_ops::ones_like_out::call(self, memory_format, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ones_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ones_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..e4a8c1b835167eb26bcebf64fa6f173694e4014d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ones_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ones(at::IntArrayRef size, ::std::optional<at::DimnameList> names, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor & ones_names_out(at::IntArrayRef size, ::std::optional<at::DimnameList> names, at::Tensor & out);
+TORCH_API at::Tensor ones(at::IntArrayRef size, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor & ones_out(at::IntArrayRef size, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/outer_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/outer_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..cc93470b3f7a409dd10cbab656f388210703987a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/outer_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 outer(const at::Tensor & self, const at::Tensor & vec2);
+TORCH_API at::Tensor & outer_out(const at::Tensor & self, const at::Tensor & vec2, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pdist.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pdist.h
new file mode 100644
index 0000000000000000000000000000000000000000..e6196d7a793700ead58fc4fbcb294253ff29f3c5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pdist.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/pdist_ops.h>
+
+namespace at {
+
+
+// aten::pdist(Tensor self, float p=2) -> Tensor
+inline at::Tensor pdist(const at::Tensor & self, double p=2) {
+    return at::_ops::pdist::call(self, p);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pdist_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pdist_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..41f0493c20651399a33e8a47c6f318230dbb82f3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pdist_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 pdist(const at::Tensor & self, double p=2);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/permute_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/permute_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..1cd3ab218ddd7952659a6ede67f0b7fd3c3ca6b4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/permute_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 permute {
+  using schema = at::Tensor (const at::Tensor &, at::IntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::permute";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "permute(Tensor(a) self, int[] dims) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, at::IntArrayRef dims);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef dims);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pixel_unshuffle_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pixel_unshuffle_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..0c4d0ee9842f4dd5b8be495eb8ee2cb9330b8a9b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pixel_unshuffle_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & pixel_unshuffle_out(const at::Tensor & self, int64_t downscale_factor, at::Tensor & out);
+TORCH_API at::Tensor pixel_unshuffle_cpu(const at::Tensor & self, int64_t downscale_factor);
+TORCH_API at::Tensor math_pixel_unshuffle(const at::Tensor & self, int64_t downscale_factor);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pixel_unshuffle_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pixel_unshuffle_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..a9d739a33ff9ed3df49ebb9b424f1612a6d23fa5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pixel_unshuffle_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 pixel_unshuffle {
+  using schema = at::Tensor (const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::pixel_unshuffle";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "pixel_unshuffle(Tensor self, int downscale_factor) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t downscale_factor);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t downscale_factor);
+};
+
+struct TORCH_API pixel_unshuffle_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::pixel_unshuffle";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "pixel_unshuffle.out(Tensor self, int downscale_factor, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, int64_t downscale_factor, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t downscale_factor, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/poisson_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/poisson_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5519e52dc4a77d3dbb1381fbf0e504b902c50c0e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/poisson_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 poisson(const at::Tensor & self, ::std::optional<at::Generator> generator=::std::nullopt);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/poisson_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/poisson_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9ca72bbc499cbe771b34828b52be1ff2c1e41740
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/poisson_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 poisson(const at::Tensor & self, ::std::optional<at::Generator> generator=::std::nullopt);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/polar_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/polar_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8b1457eac7d4c04566194d4469a257873af7a065
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/polar_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 polar {
+  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 const char* name = "aten::polar";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "polar(Tensor abs, Tensor angle) -> Tensor";
+  static at::Tensor call(const at::Tensor & abs, const at::Tensor & angle);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & abs, const at::Tensor & angle);
+};
+
+struct TORCH_API polar_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 const char* name = "aten::polar";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "polar.out(Tensor abs, Tensor angle, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & abs, const at::Tensor & angle, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & abs, const at::Tensor & angle, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/polygamma_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/polygamma_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e363b27034b5d4a352f697dab7448945a50c66f1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/polygamma_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 polygamma(int64_t n, const at::Tensor & self);
+TORCH_API at::Tensor & polygamma_out(at::Tensor & out, int64_t n, const at::Tensor & self);
+TORCH_API at::Tensor & polygamma_outf(int64_t n, const at::Tensor & self, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pow_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pow_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..581176bebf14441f05e25b9e0978e8dbbdae758a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/pow_native.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/pow_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_pow_Tensor_Tensor_out : public at::meta::structured_pow_Tensor_Tensor {
+void impl(const at::Tensor & self, const at::Tensor & exponent, const at::Tensor & out);
+};
+struct TORCH_API structured_pow_Scalar_out : public at::meta::structured_pow_Scalar {
+void impl(const at::Scalar & self, const at::Tensor & exponent, const at::Tensor & out);
+};
+struct TORCH_API structured_pow_Tensor_Scalar_out : public at::meta::structured_pow_Tensor_Scalar {
+void impl(const at::Tensor & self, const at::Scalar & exponent, const at::Tensor & out);
+};
+TORCH_API at::Tensor pow_sparse_scalar(const at::Tensor & self, const at::Scalar & exponent);
+TORCH_API at::Tensor & pow_out_sparse_scalar(const at::Tensor & self, const at::Scalar & exponent, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/prelu_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/prelu_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..da5d1c2551572b688dab3f64fca0494922110413
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/prelu_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 prelu(const at::Tensor & self, const at::Tensor & weight);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/promote_types.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/promote_types.h
new file mode 100644
index 0000000000000000000000000000000000000000..584cc6aa250000044cfbe717e22d64598f132e93
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/promote_types.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/promote_types_ops.h>
+
+namespace at {
+
+
+// aten::promote_types(ScalarType type1, ScalarType type2) -> ScalarType
+inline at::ScalarType promote_types(at::ScalarType type1, at::ScalarType type2) {
+    return at::_ops::promote_types::call(type1, type2);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/put_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/put_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..30f003f40941e5fe1ef643dfbf6a94cec7a93929
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/put_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 put(const at::Tensor & self, const at::Tensor & index, const at::Tensor & source, bool accumulate=false);
+TORCH_API at::Tensor & put_out(const at::Tensor & self, const at::Tensor & index, const at::Tensor & source, bool accumulate, at::Tensor & out);
+TORCH_API at::Tensor & put_(at::Tensor & self, const at::Tensor & index, const at::Tensor & source, bool accumulate=false);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/q_scale.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/q_scale.h
new file mode 100644
index 0000000000000000000000000000000000000000..8be4feb48832dc0c9a8d632c8d789705ec37bbe3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/q_scale.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/q_scale_ops.h>
+
+namespace at {
+
+
+// aten::q_scale(Tensor self) -> float
+inline double q_scale(const at::Tensor & self) {
+    return at::_ops::q_scale::call(self);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/q_scale_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/q_scale_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..8b5f74eb0aa5ce631a8a60a6f4e8cf3adbfa1870
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/q_scale_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 double q_scale_quant(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/qr_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/qr_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..1de67bdc4737e89542c13dcda777f584ba16dfb8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/qr_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,at::Tensor> qr(const at::Tensor & self, bool some=true);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> qr_out(at::Tensor & Q, at::Tensor & R, const at::Tensor & self, bool some=true);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> qr_outf(const at::Tensor & self, bool some, at::Tensor & Q, at::Tensor & R);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/qscheme.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/qscheme.h
new file mode 100644
index 0000000000000000000000000000000000000000..d8d12e2deb1ef5e8fc38f385e588db70dd12b591
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/qscheme.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/qscheme_ops.h>
+
+namespace at {
+
+
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/qscheme_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/qscheme_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..9073d800942b82aebe1c5ee8d387b0a91be260c2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/qscheme_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 qscheme {
+  using schema = at::QScheme (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::qscheme";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "qscheme(Tensor self) -> QScheme";
+  static at::QScheme call(const at::Tensor & self);
+  static at::QScheme redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantize_per_tensor_dynamic.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantize_per_tensor_dynamic.h
new file mode 100644
index 0000000000000000000000000000000000000000..2c0eafab7c4166c061c400bac40ad8d67c1eaecb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantize_per_tensor_dynamic.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/quantize_per_tensor_dynamic_ops.h>
+
+namespace at {
+
+
+// aten::quantize_per_tensor_dynamic(Tensor self, ScalarType dtype, bool reduce_range) -> Tensor
+inline at::Tensor quantize_per_tensor_dynamic(const at::Tensor & self, at::ScalarType dtype, bool reduce_range) {
+    return at::_ops::quantize_per_tensor_dynamic::call(self, dtype, reduce_range);
+}
+
+// aten::quantize_per_tensor_dynamic.out(Tensor self, ScalarType dtype, bool reduce_range, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & quantize_per_tensor_dynamic_out(at::Tensor & out, const at::Tensor & self, at::ScalarType dtype, bool reduce_range) {
+    return at::_ops::quantize_per_tensor_dynamic_out::call(self, dtype, reduce_range, out);
+}
+// aten::quantize_per_tensor_dynamic.out(Tensor self, ScalarType dtype, bool reduce_range, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & quantize_per_tensor_dynamic_outf(const at::Tensor & self, at::ScalarType dtype, bool reduce_range, at::Tensor & out) {
+    return at::_ops::quantize_per_tensor_dynamic_out::call(self, dtype, reduce_range, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantize_per_tensor_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantize_per_tensor_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..431bfc552555c8ff7e112a02205d228e2fb26449
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantize_per_tensor_native.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_tensor_out(const at::Tensor & self, double scale, int64_t zero_point, at::ScalarType dtype, at::Tensor & out);
+TORCH_API at::Tensor quantize_per_tensor(const at::Tensor & self, double scale, int64_t zero_point, at::ScalarType dtype);
+TORCH_API at::Tensor & quantize_per_tensor_tensor_qparams_out(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, at::ScalarType dtype, at::Tensor & out);
+TORCH_API at::Tensor quantize_per_tensor_tensor_qparams(const at::Tensor & self, const at::Tensor & scale, const at::Tensor & zero_point, at::ScalarType dtype);
+TORCH_API void quantize_per_tensor_tensors_out(at::TensorList tensors, const at::Tensor & scales, const at::Tensor & zero_points, at::ScalarType dtype, at::TensorList out);
+TORCH_API ::std::vector<at::Tensor> quantize_per_tensor_list_cpu(at::TensorList tensors, const at::Tensor & scales, const at::Tensor & zero_points, at::ScalarType dtype);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_batch_norm_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_batch_norm_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..c4e859217d3030a9f0174868026f5fb3b4db6384
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_batch_norm_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_batch_norm {
+  using schema = at::Tensor (const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const at::Tensor &, const at::Tensor &, double, double, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::quantized_batch_norm";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "quantized_batch_norm(Tensor input, Tensor? weight, Tensor? bias, Tensor mean, Tensor var, float eps, float output_scale, int output_zero_point) -> Tensor";
+  static at::Tensor call(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, const at::Tensor & mean, const at::Tensor & var, double eps, double output_scale, int64_t output_zero_point);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, const at::Tensor & mean, const at::Tensor & var, double eps, double output_scale, int64_t output_zero_point);
+};
+
+struct TORCH_API quantized_batch_norm_out {
+  using schema = at::Tensor & (const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const at::Tensor &, const at::Tensor &, double, double, int64_t, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::quantized_batch_norm";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "quantized_batch_norm.out(Tensor input, Tensor? weight, Tensor? bias, Tensor mean, Tensor var, float eps, float output_scale, int output_zero_point, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, const at::Tensor & mean, const at::Tensor & var, double eps, double output_scale, int64_t output_zero_point, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, const at::Tensor & mean, const at::Tensor & var, double eps, double output_scale, int64_t output_zero_point, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_lstm_cell.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_lstm_cell.h
new file mode 100644
index 0000000000000000000000000000000000000000..bf278acab96f97f944f45c3713d26a31edd81721
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_lstm_cell.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/quantized_lstm_cell_ops.h>
+
+namespace at {
+
+
+// aten::quantized_lstm_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, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor> quantized_lstm_cell(const at::Tensor & input, at::TensorList 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) {
+    return at::_ops::quantized_lstm_cell::call(input, hx, w_ih, w_hh, b_ih, b_hh, packed_ih, packed_hh, col_offsets_ih, col_offsets_hh, scale_ih, scale_hh, zero_point_ih, zero_point_hh);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_max_pool2d.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_max_pool2d.h
new file mode 100644
index 0000000000000000000000000000000000000000..0d78ac59e59dc61b49ace81989edb0059fed0f57
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_max_pool2d.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/quantized_max_pool2d_ops.h>
+
+namespace at {
+
+
+// aten::quantized_max_pool2d(Tensor self, int[2] kernel_size, int[2] stride=[], int[2] padding=0, int[2] dilation=1, bool ceil_mode=False) -> Tensor
+inline at::Tensor quantized_max_pool2d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, at::IntArrayRef dilation=1, bool ceil_mode=false) {
+    return at::_ops::quantized_max_pool2d::call(self, kernel_size, stride, padding, dilation, ceil_mode);
+}
+
+// aten::quantized_max_pool2d.out(Tensor self, int[2] kernel_size, int[2] stride=[], int[2] padding=0, int[2] dilation=1, bool ceil_mode=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & quantized_max_pool2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, at::IntArrayRef dilation=1, bool ceil_mode=false) {
+    return at::_ops::quantized_max_pool2d_out::call(self, kernel_size, stride, padding, dilation, ceil_mode, out);
+}
+// aten::quantized_max_pool2d.out(Tensor self, int[2] kernel_size, int[2] stride=[], int[2] padding=0, int[2] dilation=1, bool ceil_mode=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & quantized_max_pool2d_outf(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool ceil_mode, at::Tensor & out) {
+    return at::_ops::quantized_max_pool2d_out::call(self, kernel_size, stride, padding, dilation, ceil_mode, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_max_pool2d_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_max_pool2d_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b6716fd817bdf28c50c7183c3e36bcc34e52bfa7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_max_pool2d_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & quantized_max_pool2d_out(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool ceil_mode, at::Tensor & out);
+TORCH_API at::Tensor quantized_max_pool2d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, at::IntArrayRef dilation=1, bool ceil_mode=false);
+TORCH_API at::Tensor quantized_max_pool2d_cudnn(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, at::IntArrayRef dilation=1, bool ceil_mode=false);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_rnn_tanh_cell_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_rnn_tanh_cell_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..a4a57425bd04ff3a56d15705093eb9a5b89a101a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/quantized_rnn_tanh_cell_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_tanh_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 const char* name = "aten::quantized_rnn_tanh_cell";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "quantized_rnn_tanh_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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randint_like_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randint_like_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..db4ebfa16f957879d96bdfda2b57071f1714fd9e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randint_like_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,67 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 randint_like(const at::Tensor & self, int64_t high, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor randint_like(const at::Tensor & self, int64_t high, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+TORCH_API at::Tensor randint_like_symint(const at::Tensor & self, c10::SymInt high, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor randint_like_symint(const at::Tensor & self, c10::SymInt high, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+TORCH_API at::Tensor & randint_like_out(at::Tensor & out, const at::Tensor & self, int64_t high, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor & randint_like_outf(const at::Tensor & self, int64_t high, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor & randint_like_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymInt high, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor & randint_like_symint_outf(const at::Tensor & self, c10::SymInt high, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor randint_like(const at::Tensor & self, int64_t high, ::std::optional<at::Generator> generator, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor randint_like(const at::Tensor & self, int64_t high, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+TORCH_API at::Tensor randint_like_symint(const at::Tensor & self, c10::SymInt high, ::std::optional<at::Generator> generator, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor randint_like_symint(const at::Tensor & self, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+TORCH_API at::Tensor & randint_like_out(at::Tensor & out, const at::Tensor & self, int64_t high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor & randint_like_outf(const at::Tensor & self, int64_t high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor & randint_like_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor & randint_like_symint_outf(const at::Tensor & self, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor randint_like(const at::Tensor & self, const at::Tensor & high, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor randint_like(const at::Tensor & self, const at::Tensor & high, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+TORCH_API at::Tensor & randint_like_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & high, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor & randint_like_outf(const at::Tensor & self, const at::Tensor & high, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor randint_like(const at::Tensor & self, const at::Tensor & high, ::std::optional<at::Generator> generator, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor randint_like(const at::Tensor & self, const at::Tensor & high, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+TORCH_API at::Tensor & randint_like_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor & randint_like_outf(const at::Tensor & self, const at::Tensor & high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor randint_like(const at::Tensor & self, int64_t low, int64_t high, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor randint_like(const at::Tensor & self, int64_t low, int64_t high, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+TORCH_API at::Tensor randint_like_symint(const at::Tensor & self, c10::SymInt low, c10::SymInt high, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor randint_like_symint(const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+TORCH_API at::Tensor & randint_like_out(at::Tensor & out, const at::Tensor & self, int64_t low, int64_t high, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor & randint_like_outf(const at::Tensor & self, int64_t low, int64_t high, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor & randint_like_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor & randint_like_symint_outf(const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor randint_like(const at::Tensor & self, int64_t low, int64_t high, ::std::optional<at::Generator> generator, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor randint_like(const at::Tensor & self, int64_t low, int64_t high, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+TORCH_API at::Tensor randint_like_symint(const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::Generator> generator, at::TensorOptions options={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor randint_like_symint(const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+TORCH_API at::Tensor & randint_like_out(at::Tensor & out, const at::Tensor & self, int64_t low, int64_t high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor & randint_like_outf(const at::Tensor & self, int64_t low, int64_t high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+TORCH_API at::Tensor & randint_like_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+TORCH_API at::Tensor & randint_like_symint_outf(const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randint_like_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randint_like_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..f49437ffeea5ded6ff7492f546cccf4cc1f270f0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randint_like_ops.h
@@ -0,0 +1,155 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 randint_like {
+  using schema = at::Tensor (const at::Tensor &, c10::SymInt, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>, ::std::optional<at::MemoryFormat>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "randint_like(Tensor self, SymInt high, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymInt high, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt high, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+};
+
+struct TORCH_API randint_like_generator {
+  using schema = at::Tensor (const at::Tensor &, c10::SymInt, ::std::optional<at::Generator>, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>, ::std::optional<at::MemoryFormat>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "generator";
+  static constexpr const char* schema_str = "randint_like.generator(Tensor self, SymInt high, *, Generator? generator, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+};
+
+struct TORCH_API randint_like_Tensor {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>, ::std::optional<at::MemoryFormat>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "randint_like.Tensor(Tensor self, Tensor high, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & high, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & high, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+};
+
+struct TORCH_API randint_like_Tensor_generator {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, ::std::optional<at::Generator>, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>, ::std::optional<at::MemoryFormat>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "Tensor_generator";
+  static constexpr const char* schema_str = "randint_like.Tensor_generator(Tensor self, Tensor high, *, Generator? generator, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & high, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & high, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+};
+
+struct TORCH_API randint_like_low_dtype {
+  using schema = at::Tensor (const at::Tensor &, c10::SymInt, c10::SymInt, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>, ::std::optional<at::MemoryFormat>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "low_dtype";
+  static constexpr const char* schema_str = "randint_like.low_dtype(Tensor self, SymInt low, SymInt high, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+};
+
+struct TORCH_API randint_like_low_generator_dtype {
+  using schema = at::Tensor (const at::Tensor &, c10::SymInt, c10::SymInt, ::std::optional<at::Generator>, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>, ::std::optional<at::MemoryFormat>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "low_generator_dtype";
+  static constexpr const char* schema_str = "randint_like.low_generator_dtype(Tensor self, SymInt low, SymInt high, *, Generator? generator, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<at::MemoryFormat> memory_format);
+};
+
+struct TORCH_API randint_like_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymInt, ::std::optional<at::MemoryFormat>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "randint_like.out(Tensor self, SymInt high, *, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymInt high, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt high, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+};
+
+struct TORCH_API randint_like_generator_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymInt, ::std::optional<at::Generator>, ::std::optional<at::MemoryFormat>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "generator_out";
+  static constexpr const char* schema_str = "randint_like.generator_out(Tensor self, SymInt high, *, Generator? generator, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+};
+
+struct TORCH_API randint_like_Tensor_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, ::std::optional<at::MemoryFormat>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "Tensor_out";
+  static constexpr const char* schema_str = "randint_like.Tensor_out(Tensor self, Tensor high, *, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & high, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & high, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+};
+
+struct TORCH_API randint_like_Tensor_generator_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, ::std::optional<at::Generator>, ::std::optional<at::MemoryFormat>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "Tensor_generator_out";
+  static constexpr const char* schema_str = "randint_like.Tensor_generator_out(Tensor self, Tensor high, *, Generator? generator, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+};
+
+struct TORCH_API randint_like_low_dtype_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymInt, c10::SymInt, ::std::optional<at::MemoryFormat>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "low_dtype_out";
+  static constexpr const char* schema_str = "randint_like.low_dtype_out(Tensor self, SymInt low, SymInt high, *, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+};
+
+struct TORCH_API randint_like_low_generator_dtype_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymInt, c10::SymInt, ::std::optional<at::Generator>, ::std::optional<at::MemoryFormat>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::randint_like";
+  static constexpr const char* overload_name = "low_generator_dtype_out";
+  static constexpr const char* schema_str = "randint_like.low_generator_dtype_out(Tensor self, SymInt low, SymInt high, *, Generator? generator, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt low, c10::SymInt high, ::std::optional<at::Generator> generator, ::std::optional<at::MemoryFormat> memory_format, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randperm.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randperm.h
new file mode 100644
index 0000000000000000000000000000000000000000..7fd01f02b5b554e83300921ba00dd1be440d99a2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randperm.h
@@ -0,0 +1,207 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/randperm_ops.h>
+
+namespace at {
+
+
+// aten::randperm(SymInt n, *, ScalarType? dtype=long, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor randperm(int64_t n, at::TensorOptions options=at::kLong) {
+    return at::_ops::randperm::call(n, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor randperm(int64_t n, at::TensorOptions options=at::kLong) {
+    return at::_ops::randperm::call(n, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+  }
+}
+
+// aten::randperm(SymInt n, *, ScalarType? dtype=long, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor randperm(int64_t n, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::randperm::call(n, dtype, layout, device, pin_memory);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor randperm(int64_t n, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::randperm::call(n, dtype, layout, device, pin_memory);
+  }
+}
+
+// aten::randperm(SymInt n, *, ScalarType? dtype=long, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor randperm_symint(c10::SymInt n, at::TensorOptions options=at::kLong) {
+    return at::_ops::randperm::call(n, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor randperm(c10::SymInt n, at::TensorOptions options=at::kLong) {
+    return at::_ops::randperm::call(n, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+  }
+}
+
+// aten::randperm(SymInt n, *, ScalarType? dtype=long, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor randperm_symint(c10::SymInt n, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::randperm::call(n, dtype, layout, device, pin_memory);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor randperm(c10::SymInt n, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::randperm::call(n, dtype, layout, device, pin_memory);
+  }
+}
+
+// aten::randperm.generator(SymInt n, *, Generator? generator, ScalarType? dtype=long, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor randperm(int64_t n, ::std::optional<at::Generator> generator, at::TensorOptions options=at::kLong) {
+    return at::_ops::randperm_generator::call(n, generator, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor randperm(int64_t n, ::std::optional<at::Generator> generator, at::TensorOptions options=at::kLong) {
+    return at::_ops::randperm_generator::call(n, generator, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+  }
+}
+
+// aten::randperm.generator(SymInt n, *, Generator? generator, ScalarType? dtype=long, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor randperm(int64_t n, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::randperm_generator::call(n, generator, dtype, layout, device, pin_memory);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor randperm(int64_t n, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::randperm_generator::call(n, generator, dtype, layout, device, pin_memory);
+  }
+}
+
+// aten::randperm.generator(SymInt n, *, Generator? generator, ScalarType? dtype=long, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor randperm_symint(c10::SymInt n, ::std::optional<at::Generator> generator, at::TensorOptions options=at::kLong) {
+    return at::_ops::randperm_generator::call(n, generator, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor randperm(c10::SymInt n, ::std::optional<at::Generator> generator, at::TensorOptions options=at::kLong) {
+    return at::_ops::randperm_generator::call(n, generator, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+  }
+}
+
+// aten::randperm.generator(SymInt n, *, Generator? generator, ScalarType? dtype=long, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor randperm_symint(c10::SymInt n, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::randperm_generator::call(n, generator, dtype, layout, device, pin_memory);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor randperm(c10::SymInt n, ::std::optional<at::Generator> generator, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::randperm_generator::call(n, generator, dtype, layout, device, pin_memory);
+  }
+}
+
+// aten::randperm.out(SymInt n, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & randperm_out(at::Tensor & out, int64_t n) {
+    return at::_ops::randperm_out::call(n, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & randperm_out(at::Tensor & out, int64_t n) {
+    return at::_ops::randperm_out::call(n, out);
+  }
+}
+
+// aten::randperm.out(SymInt n, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & randperm_outf(int64_t n, at::Tensor & out) {
+    return at::_ops::randperm_out::call(n, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & randperm_outf(int64_t n, at::Tensor & out) {
+    return at::_ops::randperm_out::call(n, out);
+  }
+}
+
+// aten::randperm.out(SymInt n, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & randperm_symint_out(at::Tensor & out, c10::SymInt n) {
+    return at::_ops::randperm_out::call(n, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & randperm_out(at::Tensor & out, c10::SymInt n) {
+    return at::_ops::randperm_out::call(n, out);
+  }
+}
+
+// aten::randperm.out(SymInt n, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & randperm_symint_outf(c10::SymInt n, at::Tensor & out) {
+    return at::_ops::randperm_out::call(n, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & randperm_outf(c10::SymInt n, at::Tensor & out) {
+    return at::_ops::randperm_out::call(n, out);
+  }
+}
+
+// aten::randperm.generator_out(SymInt n, *, Generator? generator, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & randperm_out(at::Tensor & out, int64_t n, ::std::optional<at::Generator> generator) {
+    return at::_ops::randperm_generator_out::call(n, generator, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & randperm_out(at::Tensor & out, int64_t n, ::std::optional<at::Generator> generator) {
+    return at::_ops::randperm_generator_out::call(n, generator, out);
+  }
+}
+
+// aten::randperm.generator_out(SymInt n, *, Generator? generator, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & randperm_outf(int64_t n, ::std::optional<at::Generator> generator, at::Tensor & out) {
+    return at::_ops::randperm_generator_out::call(n, generator, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & randperm_outf(int64_t n, ::std::optional<at::Generator> generator, at::Tensor & out) {
+    return at::_ops::randperm_generator_out::call(n, generator, out);
+  }
+}
+
+// aten::randperm.generator_out(SymInt n, *, Generator? generator, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & randperm_symint_out(at::Tensor & out, c10::SymInt n, ::std::optional<at::Generator> generator) {
+    return at::_ops::randperm_generator_out::call(n, generator, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & randperm_out(at::Tensor & out, c10::SymInt n, ::std::optional<at::Generator> generator) {
+    return at::_ops::randperm_generator_out::call(n, generator, out);
+  }
+}
+
+// aten::randperm.generator_out(SymInt n, *, Generator? generator, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & randperm_symint_outf(c10::SymInt n, ::std::optional<at::Generator> generator, at::Tensor & out) {
+    return at::_ops::randperm_generator_out::call(n, generator, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & randperm_outf(c10::SymInt n, ::std::optional<at::Generator> generator, at::Tensor & out) {
+    return at::_ops::randperm_generator_out::call(n, generator, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randperm_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randperm_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..96aabdd08b39a05a5c053a7ccb406c36426bb1e6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/randperm_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/range_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/range_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..a4b1e65a6f4ab385756bafbb1f4b783f24255924
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/range_native.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 range(const at::Scalar & start, const at::Scalar & end, const at::Scalar & step=1, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor & range_out(const at::Scalar & start, const at::Scalar & end, const at::Scalar & step, at::Tensor & out);
+TORCH_API at::Tensor & range_cuda_out(const at::Scalar & start, const at::Scalar & end, const at::Scalar & step, at::Tensor & out);
+TORCH_API at::Tensor range(const at::Scalar & start, const at::Scalar & end, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor & range_out_no_step(const at::Scalar & start, const at::Scalar & end, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/range_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/range_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..62d8337758f60872af12a0a1571a39f801bc9b44
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/range_ops.h
@@ -0,0 +1,67 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 range_step {
+  using schema = at::Tensor (const at::Scalar &, const at::Scalar &, const at::Scalar &, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::range";
+  static constexpr const char* overload_name = "step";
+  static constexpr const char* schema_str = "range.step(Scalar start, Scalar end, Scalar step=1, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor";
+  static at::Tensor call(const at::Scalar & start, const at::Scalar & end, const at::Scalar & step, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & start, const at::Scalar & end, const at::Scalar & step, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+struct TORCH_API range {
+  using schema = at::Tensor (const at::Scalar &, const at::Scalar &, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::range";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "range(Scalar start, Scalar end, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor";
+  static at::Tensor call(const at::Scalar & start, const at::Scalar & end, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & start, const at::Scalar & end, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+struct TORCH_API range_out_ {
+  using schema = at::Tensor & (const at::Scalar &, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::range";
+  static constexpr const char* overload_name = "out_";
+  static constexpr const char* schema_str = "range.out_(Scalar start, Scalar end, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Scalar & start, const at::Scalar & end, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & start, const at::Scalar & end, at::Tensor & out);
+};
+
+struct TORCH_API range_out {
+  using schema = at::Tensor & (const at::Scalar &, const at::Scalar &, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::range";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "range.out(Scalar start, Scalar end, Scalar step=1, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Scalar & start, const at::Scalar & end, const at::Scalar & step, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & start, const at::Scalar & end, const at::Scalar & step, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ravel_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ravel_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..439ae73db9b3aa9b14065f7e1584faf3f6ac9a35
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/ravel_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 ravel {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::ravel";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "ravel(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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/real_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/real_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..e78f4441c2b37e087f699f00445e2644405a2588
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/real_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 real {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::real";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "real(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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reciprocal_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reciprocal_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..52af20b1c6299f7f7326b9c8de5325a2fbafe183
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reciprocal_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 reciprocal(const at::Tensor & self);
+TORCH_API at::Tensor & reciprocal_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & reciprocal_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & reciprocal_(at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/refine_names_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/refine_names_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..15d13b42f65bc54988404e572389c02a0666fd51
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/refine_names_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 refine_names(const at::Tensor & self, at::DimnameList names);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad1d_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad1d_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..bb77b4c04f708ca1d1449d7ec31013d240bf6c61
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad1d_cpu_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 reflection_pad1d(const at::Tensor & self, at::IntArrayRef padding);
+TORCH_API at::Tensor reflection_pad1d_symint(const at::Tensor & self, c10::SymIntArrayRef padding);
+TORCH_API at::Tensor & reflection_pad1d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef padding);
+TORCH_API at::Tensor & reflection_pad1d_outf(const at::Tensor & self, at::IntArrayRef padding, at::Tensor & out);
+TORCH_API at::Tensor & reflection_pad1d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef padding);
+TORCH_API at::Tensor & reflection_pad1d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef padding, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad1d_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad1d_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..8dcbdcebf6f19b73fb333df05190415ff8b60fc0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad1d_native.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/reflection_pad1d_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_reflection_pad1d_out_cpu : public at::meta::structured_reflection_pad1d {
+void impl(const at::Tensor & self, at::ArrayRef<int64_t> padding, const at::Tensor & out);
+};
+struct TORCH_API structured_reflection_pad1d_out_cuda : public at::meta::structured_reflection_pad1d {
+void impl(const at::Tensor & self, at::ArrayRef<int64_t> padding, const at::Tensor & out);
+};
+TORCH_API at::Tensor & reflection_pad1d_out_quantized_cpu(const at::Tensor & self, at::IntArrayRef padding, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad2d.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad2d.h
new file mode 100644
index 0000000000000000000000000000000000000000..cf8da2a8da9600e5208af519309dc9e367fb2a62
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad2d.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/reflection_pad2d_ops.h>
+
+namespace at {
+
+
+// aten::reflection_pad2d.out(Tensor self, SymInt[4] padding, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & reflection_pad2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef padding) {
+    return at::_ops::reflection_pad2d_out::call(self, c10::fromIntArrayRefSlow(padding), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & reflection_pad2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef padding) {
+    return at::_ops::reflection_pad2d_out::call(self, c10::fromIntArrayRefSlow(padding), out);
+  }
+}
+
+// aten::reflection_pad2d.out(Tensor self, SymInt[4] padding, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & reflection_pad2d_outf(const at::Tensor & self, at::IntArrayRef padding, at::Tensor & out) {
+    return at::_ops::reflection_pad2d_out::call(self, c10::fromIntArrayRefSlow(padding), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & reflection_pad2d_outf(const at::Tensor & self, at::IntArrayRef padding, at::Tensor & out) {
+    return at::_ops::reflection_pad2d_out::call(self, c10::fromIntArrayRefSlow(padding), out);
+  }
+}
+
+// aten::reflection_pad2d.out(Tensor self, SymInt[4] padding, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & reflection_pad2d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef padding) {
+    return at::_ops::reflection_pad2d_out::call(self, padding, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & reflection_pad2d_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef padding) {
+    return at::_ops::reflection_pad2d_out::call(self, padding, out);
+  }
+}
+
+// aten::reflection_pad2d.out(Tensor self, SymInt[4] padding, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & reflection_pad2d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef padding, at::Tensor & out) {
+    return at::_ops::reflection_pad2d_out::call(self, padding, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & reflection_pad2d_outf(const at::Tensor & self, c10::SymIntArrayRef padding, at::Tensor & out) {
+    return at::_ops::reflection_pad2d_out::call(self, padding, out);
+  }
+}
+
+// aten::reflection_pad2d(Tensor self, SymInt[4] padding) -> Tensor
+inline at::Tensor reflection_pad2d(const at::Tensor & self, at::IntArrayRef padding) {
+    return at::_ops::reflection_pad2d::call(self, c10::fromIntArrayRefSlow(padding));
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor reflection_pad2d(const at::Tensor & self, at::IntArrayRef padding) {
+    return at::_ops::reflection_pad2d::call(self, c10::fromIntArrayRefSlow(padding));
+  }
+}
+
+// aten::reflection_pad2d(Tensor self, SymInt[4] padding) -> Tensor
+inline at::Tensor reflection_pad2d_symint(const at::Tensor & self, c10::SymIntArrayRef padding) {
+    return at::_ops::reflection_pad2d::call(self, padding);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor reflection_pad2d(const at::Tensor & self, c10::SymIntArrayRef padding) {
+    return at::_ops::reflection_pad2d::call(self, padding);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad2d_backward_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad2d_backward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..88de6008e364a83674c442b019d1f36093ec8d9e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad2d_backward_cuda_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 reflection_pad2d_backward(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef padding);
+TORCH_API at::Tensor reflection_pad2d_backward_symint(const at::Tensor & grad_output, const at::Tensor & self, c10::SymIntArrayRef padding);
+TORCH_API at::Tensor & reflection_pad2d_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef padding);
+TORCH_API at::Tensor & reflection_pad2d_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, at::IntArrayRef padding, at::Tensor & grad_input);
+TORCH_API at::Tensor & reflection_pad2d_backward_symint_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, c10::SymIntArrayRef padding);
+TORCH_API at::Tensor & reflection_pad2d_backward_symint_outf(const at::Tensor & grad_output, const at::Tensor & self, c10::SymIntArrayRef padding, at::Tensor & grad_input);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad3d_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad3d_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..c2fd9f0d0a960a85eb1bf8094eac525f6ab07780
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reflection_pad3d_backward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 reflection_pad3d_backward_grad_input {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::reflection_pad3d_backward";
+  static constexpr const char* overload_name = "grad_input";
+  static constexpr const char* schema_str = "reflection_pad3d_backward.grad_input(Tensor grad_output, Tensor self, SymInt[6] padding, *, Tensor(a!) grad_input) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & grad_output, const at::Tensor & self, c10::SymIntArrayRef padding, at::Tensor & grad_input);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self, c10::SymIntArrayRef padding, at::Tensor & grad_input);
+};
+
+struct TORCH_API reflection_pad3d_backward {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::reflection_pad3d_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "reflection_pad3d_backward(Tensor grad_output, Tensor self, SymInt[6] padding) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad_output, const at::Tensor & self, c10::SymIntArrayRef padding);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self, c10::SymIntArrayRef padding);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/remainder_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/remainder_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..336fb4560d27fbd8e9a4ccf75c400098c69b27be
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/remainder_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 remainder(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & remainder_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/renorm_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/renorm_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..d06391f123e418772bbc249fbe6f702691b18b97
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/renorm_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/renorm_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_renorm_out : public at::meta::structured_renorm {
+void impl(const at::Tensor & self, const at::Scalar & p, int64_t dim, const at::Scalar & maxnorm, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..54829437eab67026aeb40fef118ae2aa003ed8d0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 repeat(const at::Tensor & self, at::IntArrayRef repeats);
+TORCH_API at::Tensor repeat_symint(const at::Tensor & self, c10::SymIntArrayRef repeats);
+TORCH_API at::Tensor & repeat_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef repeats);
+TORCH_API at::Tensor & repeat_outf(const at::Tensor & self, at::IntArrayRef repeats, at::Tensor & out);
+TORCH_API at::Tensor & repeat_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef repeats);
+TORCH_API at::Tensor & repeat_symint_outf(const at::Tensor & self, c10::SymIntArrayRef repeats, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_interleave_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_interleave_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c6b8295bbc94117cdc8d820d69e63e5bf3957684
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_interleave_cpu_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 repeat_interleave(const at::Tensor & repeats, ::std::optional<int64_t> output_size=::std::nullopt);
+TORCH_API at::Tensor repeat_interleave_symint(const at::Tensor & repeats, ::std::optional<c10::SymInt> output_size=::std::nullopt);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_interleave_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_interleave_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..a994ac0ab4289c88095ea0e231e24cd7302af194
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_interleave_ops.h
@@ -0,0 +1,67 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 repeat_interleave_Tensor {
+  using schema = at::Tensor (const at::Tensor &, ::std::optional<c10::SymInt>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::repeat_interleave";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "repeat_interleave.Tensor(Tensor repeats, *, SymInt? output_size=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & repeats, ::std::optional<c10::SymInt> output_size);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & repeats, ::std::optional<c10::SymInt> output_size);
+};
+
+struct TORCH_API repeat_interleave_self_Tensor {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, ::std::optional<int64_t>, ::std::optional<c10::SymInt>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::repeat_interleave";
+  static constexpr const char* overload_name = "self_Tensor";
+  static constexpr const char* schema_str = "repeat_interleave.self_Tensor(Tensor self, Tensor repeats, int? dim=None, *, SymInt? output_size=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & repeats, ::std::optional<int64_t> dim, ::std::optional<c10::SymInt> output_size);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & repeats, ::std::optional<int64_t> dim, ::std::optional<c10::SymInt> output_size);
+};
+
+struct TORCH_API repeat_interleave_self_int {
+  using schema = at::Tensor (const at::Tensor &, c10::SymInt, ::std::optional<int64_t>, ::std::optional<c10::SymInt>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::repeat_interleave";
+  static constexpr const char* overload_name = "self_int";
+  static constexpr const char* schema_str = "repeat_interleave.self_int(Tensor self, SymInt repeats, int? dim=None, *, SymInt? output_size=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymInt repeats, ::std::optional<int64_t> dim, ::std::optional<c10::SymInt> output_size);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt repeats, ::std::optional<int64_t> dim, ::std::optional<c10::SymInt> output_size);
+};
+
+struct TORCH_API repeat_interleave_Tensor_out {
+  using schema = at::Tensor & (const at::Tensor &, ::std::optional<c10::SymInt>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::repeat_interleave";
+  static constexpr const char* overload_name = "Tensor_out";
+  static constexpr const char* schema_str = "repeat_interleave.Tensor_out(Tensor repeats, *, SymInt? output_size=None, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & repeats, ::std::optional<c10::SymInt> output_size, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & repeats, ::std::optional<c10::SymInt> output_size, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..407f761d7bfd46cb33f499215572847fb8952baf
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/repeat_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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(const at::Tensor & self, at::IntArrayRef repeats);
+TORCH_API at::Tensor & repeat_out_symint(const at::Tensor & self, c10::SymIntArrayRef repeats, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad1d_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad1d_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e004159a19b40534d5a9efd103004f1d66dff79a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad1d_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 replication_pad1d(const at::Tensor & self, at::IntArrayRef padding);
+TORCH_API at::Tensor replication_pad1d_symint(const at::Tensor & self, c10::SymIntArrayRef padding);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad1d_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad1d_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..5dce80519c41cebd68d282550f90ae56249435ac
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad1d_native.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/replication_pad1d_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_replication_pad1d_out_cpu : public at::meta::structured_replication_pad1d {
+void impl(const at::Tensor & self, at::ArrayRef<int64_t> padding, const at::Tensor & out);
+};
+struct TORCH_API structured_replication_pad1d_out_cuda : public at::meta::structured_replication_pad1d {
+void impl(const at::Tensor & self, at::ArrayRef<int64_t> padding, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad2d_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad2d_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..ed646d2a0c5fc5ac0b37ca05b56a58324094f1f1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad2d_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_replication_pad2d : public at::impl::MetaBase {
+
+
+    void meta(const at::Tensor & self, at::ArrayRef<int64_t> padding);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad2d_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad2d_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..13582ad7eb7b0ee19c0c23cca7f88f415b622e44
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad2d_meta_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 replication_pad2d(const at::Tensor & self, at::IntArrayRef padding);
+TORCH_API at::Tensor replication_pad2d_symint(const at::Tensor & self, c10::SymIntArrayRef padding);
+TORCH_API at::Tensor & replication_pad2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef padding);
+TORCH_API at::Tensor & replication_pad2d_outf(const at::Tensor & self, at::IntArrayRef padding, at::Tensor & out);
+TORCH_API at::Tensor & replication_pad2d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef padding);
+TORCH_API at::Tensor & replication_pad2d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef padding, at::Tensor & out);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad3d_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad3d_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..b138004f2f04b11aa560045579720b1f3f10c91f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad3d_backward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 replication_pad3d_backward_grad_input {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::replication_pad3d_backward";
+  static constexpr const char* overload_name = "grad_input";
+  static constexpr const char* schema_str = "replication_pad3d_backward.grad_input(Tensor grad_output, Tensor self, SymInt[6] padding, *, Tensor(a!) grad_input) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & grad_output, const at::Tensor & self, c10::SymIntArrayRef padding, at::Tensor & grad_input);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self, c10::SymIntArrayRef padding, at::Tensor & grad_input);
+};
+
+struct TORCH_API replication_pad3d_backward {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::replication_pad3d_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "replication_pad3d_backward(Tensor grad_output, Tensor self, SymInt[6] padding) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad_output, const at::Tensor & self, c10::SymIntArrayRef padding);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self, c10::SymIntArrayRef padding);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad3d_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad3d_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f1c9cd733c92af789370fc14bf4620fd4738679b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad3d_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 replication_pad3d(const at::Tensor & self, at::IntArrayRef padding);
+TORCH_API at::Tensor replication_pad3d_symint(const at::Tensor & self, c10::SymIntArrayRef padding);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad3d_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad3d_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4bac2e7cd5cce6493d8428d9e60606ae16a5c663
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/replication_pad3d_cpu_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 replication_pad3d(const at::Tensor & self, at::IntArrayRef padding);
+TORCH_API at::Tensor replication_pad3d_symint(const at::Tensor & self, c10::SymIntArrayRef padding);
+TORCH_API at::Tensor & replication_pad3d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef padding);
+TORCH_API at::Tensor & replication_pad3d_outf(const at::Tensor & self, at::IntArrayRef padding, at::Tensor & out);
+TORCH_API at::Tensor & replication_pad3d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef padding);
+TORCH_API at::Tensor & replication_pad3d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef padding, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reshape.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reshape.h
new file mode 100644
index 0000000000000000000000000000000000000000..9874cc5632dbcc21a58a393536a404bffab392d4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reshape.h
@@ -0,0 +1,53 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/reshape_ops.h>
+
+namespace at {
+
+
+// aten::reshape(Tensor(a) self, SymInt[] shape) -> Tensor(a)
+inline at::Tensor reshape(const at::Tensor & self, at::IntArrayRef shape) {
+    return at::_ops::reshape::call(self, c10::fromIntArrayRefSlow(shape));
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor reshape(const at::Tensor & self, at::IntArrayRef shape) {
+    return at::_ops::reshape::call(self, c10::fromIntArrayRefSlow(shape));
+  }
+}
+
+// aten::reshape(Tensor(a) self, SymInt[] shape) -> Tensor(a)
+inline at::Tensor reshape_symint(const at::Tensor & self, c10::SymIntArrayRef shape) {
+    return at::_ops::reshape::call(self, shape);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor reshape(const at::Tensor & self, c10::SymIntArrayRef shape) {
+    return at::_ops::reshape::call(self, shape);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reshape_as_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reshape_as_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..032e3bfc0e9bfc3ac1752b24cff3f4ff2bd1082e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/reshape_as_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 reshape_as(const at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/resize_as_sparse_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/resize_as_sparse_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ab29330f883f808450ca280f684991c5dbd5f9c4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/resize_as_sparse_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 resize_as_sparse(const at::Tensor & self, const at::Tensor & the_template);
+TORCH_API const at::Tensor & resize_as_sparse_out(const at::Tensor & out, const at::Tensor & self, const at::Tensor & the_template);
+TORCH_API const at::Tensor & resize_as_sparse_outf(const at::Tensor & self, const at::Tensor & the_template, const at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/resize_as_sparse_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/resize_as_sparse_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9e9dabc5e628e795136360bd6c34b7f51a1e3074
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/resize_as_sparse_meta_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 const at::Tensor & resize_as_sparse_(const at::Tensor & self, const at::Tensor & the_template);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/resolve_neg_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/resolve_neg_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..931615b23c7f36ae28a98f7f3defdb69bca3fb1e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/resolve_neg_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 resolve_neg {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::resolve_neg";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "resolve_neg(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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/result_type.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/result_type.h
new file mode 100644
index 0000000000000000000000000000000000000000..56a5c023f724d748b3045762ce86c463d6dea1fe
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/result_type.h
@@ -0,0 +1,51 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/result_type_ops.h>
+
+namespace at {
+
+
+// aten::result_type.Tensor(Tensor tensor, Tensor other) -> ScalarType
+inline at::ScalarType result_type(const at::Tensor & tensor, const at::Tensor & other) {
+    return at::_ops::result_type_Tensor::call(tensor, other);
+}
+
+// aten::result_type.Scalar(Tensor tensor, Scalar other) -> ScalarType
+inline at::ScalarType result_type(const at::Tensor & tensor, const at::Scalar & other) {
+    return at::_ops::result_type_Scalar::call(tensor, other);
+}
+
+// aten::result_type.Scalar_Tensor(Scalar scalar, Tensor tensor) -> ScalarType
+inline at::ScalarType result_type(const at::Scalar & scalar, const at::Tensor & tensor) {
+    return at::_ops::result_type_Scalar_Tensor::call(scalar, tensor);
+}
+
+// aten::result_type.Scalar_Scalar(Scalar scalar1, Scalar scalar2) -> ScalarType
+inline at::ScalarType result_type(const at::Scalar & scalar1, const at::Scalar & scalar2) {
+    return at::_ops::result_type_Scalar_Scalar::call(scalar1, scalar2);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/result_type_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/result_type_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..4d0cc2433367d85723dff487e61ca7f18cf9a4b5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/result_type_ops.h
@@ -0,0 +1,67 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 result_type_Tensor {
+  using schema = at::ScalarType (const at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::result_type";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "result_type.Tensor(Tensor tensor, Tensor other) -> ScalarType";
+  static at::ScalarType call(const at::Tensor & tensor, const at::Tensor & other);
+  static at::ScalarType redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & tensor, const at::Tensor & other);
+};
+
+struct TORCH_API result_type_Scalar {
+  using schema = at::ScalarType (const at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::result_type";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "result_type.Scalar(Tensor tensor, Scalar other) -> ScalarType";
+  static at::ScalarType call(const at::Tensor & tensor, const at::Scalar & other);
+  static at::ScalarType redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & tensor, const at::Scalar & other);
+};
+
+struct TORCH_API result_type_Scalar_Tensor {
+  using schema = at::ScalarType (const at::Scalar &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::result_type";
+  static constexpr const char* overload_name = "Scalar_Tensor";
+  static constexpr const char* schema_str = "result_type.Scalar_Tensor(Scalar scalar, Tensor tensor) -> ScalarType";
+  static at::ScalarType call(const at::Scalar & scalar, const at::Tensor & tensor);
+  static at::ScalarType redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & scalar, const at::Tensor & tensor);
+};
+
+struct TORCH_API result_type_Scalar_Scalar {
+  using schema = at::ScalarType (const at::Scalar &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::result_type";
+  static constexpr const char* overload_name = "Scalar_Scalar";
+  static constexpr const char* schema_str = "result_type.Scalar_Scalar(Scalar scalar1, Scalar scalar2) -> ScalarType";
+  static at::ScalarType call(const at::Scalar & scalar1, const at::Scalar & scalar2);
+  static at::ScalarType redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & scalar1, const at::Scalar & scalar2);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/retains_grad_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/retains_grad_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..750133e0173baddd34b645562b1d62d0bb4f834b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/retains_grad_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API bool retains_grad(const at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rms_norm.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rms_norm.h
new file mode 100644
index 0000000000000000000000000000000000000000..d01a5c951ae79be1bba49f483561011a88567033
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rms_norm.h
@@ -0,0 +1,53 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/rms_norm_ops.h>
+
+namespace at {
+
+
+// aten::rms_norm(Tensor input, SymInt[] normalized_shape, Tensor? weight=None, float? eps=None) -> Tensor
+inline at::Tensor rms_norm(const at::Tensor & input, at::IntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight={}, ::std::optional<double> eps=::std::nullopt) {
+    return at::_ops::rms_norm::call(input, c10::fromIntArrayRefSlow(normalized_shape), weight, eps);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor rms_norm(const at::Tensor & input, at::IntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight={}, ::std::optional<double> eps=::std::nullopt) {
+    return at::_ops::rms_norm::call(input, c10::fromIntArrayRefSlow(normalized_shape), weight, eps);
+  }
+}
+
+// aten::rms_norm(Tensor input, SymInt[] normalized_shape, Tensor? weight=None, float? eps=None) -> Tensor
+inline at::Tensor rms_norm_symint(const at::Tensor & input, c10::SymIntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight={}, ::std::optional<double> eps=::std::nullopt) {
+    return at::_ops::rms_norm::call(input, normalized_shape, weight, eps);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor rms_norm(const at::Tensor & input, c10::SymIntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight={}, ::std::optional<double> eps=::std::nullopt) {
+    return at::_ops::rms_norm::call(input, normalized_shape, weight, eps);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/roll.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/roll.h
new file mode 100644
index 0000000000000000000000000000000000000000..c5381172983926e85555d95d43353e604df81f03
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/roll.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/roll_ops.h>
+
+namespace at {
+
+
+// aten::roll(Tensor self, SymInt[1] shifts, int[1] dims=[]) -> Tensor
+inline at::Tensor roll(const at::Tensor & self, at::IntArrayRef shifts, at::IntArrayRef dims={}) {
+    return at::_ops::roll::call(self, c10::fromIntArrayRefSlow(shifts), dims);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor roll(const at::Tensor & self, at::IntArrayRef shifts, at::IntArrayRef dims={}) {
+    return at::_ops::roll::call(self, c10::fromIntArrayRefSlow(shifts), dims);
+  }
+}
+
+// aten::roll(Tensor self, SymInt[1] shifts, int[1] dims=[]) -> Tensor
+inline at::Tensor roll_symint(const at::Tensor & self, c10::SymIntArrayRef shifts, at::IntArrayRef dims={}) {
+    return at::_ops::roll::call(self, shifts, dims);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor roll(const at::Tensor & self, c10::SymIntArrayRef shifts, at::IntArrayRef dims={}) {
+    return at::_ops::roll::call(self, shifts, dims);
+  }
+}
+
+// aten::roll.out(Tensor self, SymInt[1] shifts, int[1] dims=[], *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & roll_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef shifts, at::IntArrayRef dims={}) {
+    return at::_ops::roll_out::call(self, c10::fromIntArrayRefSlow(shifts), dims, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & roll_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef shifts, at::IntArrayRef dims={}) {
+    return at::_ops::roll_out::call(self, c10::fromIntArrayRefSlow(shifts), dims, out);
+  }
+}
+
+// aten::roll.out(Tensor self, SymInt[1] shifts, int[1] dims=[], *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & roll_outf(const at::Tensor & self, at::IntArrayRef shifts, at::IntArrayRef dims, at::Tensor & out) {
+    return at::_ops::roll_out::call(self, c10::fromIntArrayRefSlow(shifts), dims, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & roll_outf(const at::Tensor & self, at::IntArrayRef shifts, at::IntArrayRef dims, at::Tensor & out) {
+    return at::_ops::roll_out::call(self, c10::fromIntArrayRefSlow(shifts), dims, out);
+  }
+}
+
+// aten::roll.out(Tensor self, SymInt[1] shifts, int[1] dims=[], *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & roll_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef shifts, at::IntArrayRef dims={}) {
+    return at::_ops::roll_out::call(self, shifts, dims, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & roll_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef shifts, at::IntArrayRef dims={}) {
+    return at::_ops::roll_out::call(self, shifts, dims, out);
+  }
+}
+
+// aten::roll.out(Tensor self, SymInt[1] shifts, int[1] dims=[], *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & roll_symint_outf(const at::Tensor & self, c10::SymIntArrayRef shifts, at::IntArrayRef dims, at::Tensor & out) {
+    return at::_ops::roll_out::call(self, shifts, dims, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & roll_outf(const at::Tensor & self, c10::SymIntArrayRef shifts, at::IntArrayRef dims, at::Tensor & out) {
+    return at::_ops::roll_out::call(self, shifts, dims, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/roll_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/roll_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..2d28ec33460d095e3c9a7e2c54d4c6f37bd7c148
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/roll_cuda_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 roll(const at::Tensor & self, at::IntArrayRef shifts, at::IntArrayRef dims={});
+TORCH_API at::Tensor roll_symint(const at::Tensor & self, c10::SymIntArrayRef shifts, at::IntArrayRef dims={});
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/roll_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/roll_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..8077c42cd2761b83c955866bc4a0f6ccd3ddd5d9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/roll_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 roll {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef, at::IntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::roll";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "roll(Tensor self, SymInt[1] shifts, int[1] dims=[]) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymIntArrayRef shifts, at::IntArrayRef dims);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef shifts, at::IntArrayRef dims);
+};
+
+struct TORCH_API roll_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymIntArrayRef, at::IntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::roll";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "roll.out(Tensor self, SymInt[1] shifts, int[1] dims=[], *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymIntArrayRef shifts, at::IntArrayRef dims, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef shifts, at::IntArrayRef dims, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rot90_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rot90_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..7281a7000203abc41412759526e1185f1fa3655c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rot90_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 rot90(const at::Tensor & self, int64_t k=1, at::IntArrayRef dims={0,1});
+TORCH_API at::Tensor & rot90_out(const at::Tensor & self, int64_t k, at::IntArrayRef dims, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/round_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/round_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..c3548bdb3808c0f571de20983c4e10c425b6579b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/round_native.h
@@ -0,0 +1,37 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/round_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_round_out : public at::meta::structured_round {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+TORCH_API at::Tensor round_sparse(const at::Tensor & self);
+TORCH_API at::Tensor & round_sparse_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & round_sparse_(at::Tensor & self);
+TORCH_API at::Tensor round_sparse_csr(const at::Tensor & self);
+TORCH_API at::Tensor & round_sparse_csr_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & round_sparse_csr_(at::Tensor & self);
+struct TORCH_API structured_round_decimals_out : public at::meta::structured_round_decimals {
+void impl(const at::Tensor & self, int64_t decimals, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/row_indices_copy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/row_indices_copy.h
new file mode 100644
index 0000000000000000000000000000000000000000..3e49564495312d4966ba276fae13c79b1187becd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/row_indices_copy.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/row_indices_copy_ops.h>
+
+namespace at {
+
+
+// aten::row_indices_copy(Tensor self) -> Tensor
+inline at::Tensor row_indices_copy(const at::Tensor & self) {
+    return at::_ops::row_indices_copy::call(self);
+}
+
+// aten::row_indices_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & row_indices_copy_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::row_indices_copy_out::call(self, out);
+}
+// aten::row_indices_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & row_indices_copy_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::row_indices_copy_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/row_stack_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/row_stack_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f711e2cd74fbb5c686f48ab98cf4ab2a9f359ef1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/row_stack_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rrelu_with_noise_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rrelu_with_noise_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e7a423dd3a2dc9dce1ac861bc37755cb696a3d01
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rrelu_with_noise_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & rrelu_with_noise_out(at::Tensor & out, const at::Tensor & self, at::Tensor & noise, const at::Scalar & lower=0.125, const at::Scalar & upper=0.3333333333333333, bool training=false, ::std::optional<at::Generator> generator=::std::nullopt);
+TORCH_API at::Tensor & rrelu_with_noise_outf(const at::Tensor & self, at::Tensor & noise, const at::Scalar & lower, const at::Scalar & upper, bool training, ::std::optional<at::Generator> generator, at::Tensor & out);
+TORCH_API at::Tensor & rrelu_with_noise_(at::Tensor & self, at::Tensor & noise, const at::Scalar & lower=0.125, const at::Scalar & upper=0.3333333333333333, bool training=false, ::std::optional<at::Generator> generator=::std::nullopt);
+TORCH_API at::Tensor rrelu_with_noise(const at::Tensor & self, at::Tensor & noise, const at::Scalar & lower=0.125, const at::Scalar & upper=0.3333333333333333, bool training=false, ::std::optional<at::Generator> generator=::std::nullopt);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rsqrt_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rsqrt_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..9fe6dbaf35724793b7783b040aa9333df7adfa23
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/rsqrt_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/rsqrt_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_rsqrt_out : public at::meta::structured_rsqrt {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scalar_tensor_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scalar_tensor_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..1f3d31c7352494a8565e7ae8f6f8896749b5541b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scalar_tensor_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 scalar_tensor {
+  using schema = at::Tensor (const at::Scalar &, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scalar_tensor";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "scalar_tensor(Scalar s, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor";
+  static at::Tensor call(const at::Scalar & s, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & s, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+struct TORCH_API scalar_tensor_out {
+  using schema = at::Tensor & (const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scalar_tensor";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "scalar_tensor.out(Scalar s, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Scalar & s, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & s, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..473be59a630208ef0d67699bf9e7826ff4858b22
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 scatter(const at::Tensor & self, at::Dimname dim, const at::Tensor & index, const at::Tensor & src);
+TORCH_API at::Tensor scatter(const at::Tensor & self, at::Dimname dim, const at::Tensor & index, const at::Scalar & value);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b1fe098221c696101efe81e13fbaff51c0d16a03
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_cuda_dispatch.h
@@ -0,0 +1,43 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 scatter(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src);
+TORCH_API at::Tensor & scatter_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src);
+TORCH_API at::Tensor & scatter_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, at::Tensor & out);
+TORCH_API at::Tensor & scatter_(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src);
+TORCH_API at::Tensor scatter(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value);
+TORCH_API at::Tensor & scatter_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value);
+TORCH_API at::Tensor & scatter_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, at::Tensor & out);
+TORCH_API at::Tensor & scatter_(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value);
+TORCH_API at::Tensor scatter(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce);
+TORCH_API at::Tensor & scatter_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce);
+TORCH_API at::Tensor & scatter_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, at::Tensor & out);
+TORCH_API at::Tensor & scatter_(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce);
+TORCH_API at::Tensor scatter(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, c10::string_view reduce);
+TORCH_API at::Tensor & scatter_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, c10::string_view reduce);
+TORCH_API at::Tensor & scatter_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, c10::string_view reduce, at::Tensor & out);
+TORCH_API at::Tensor & scatter_(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, c10::string_view reduce);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..49f977d86d035666fb75eed0e671c9dedf45c537
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_ops.h
@@ -0,0 +1,177 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 scatter_src {
+  using schema = at::Tensor (const at::Tensor &, int64_t, const at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter";
+  static constexpr const char* overload_name = "src";
+  static constexpr const char* schema_str = "scatter.src(Tensor self, int dim, Tensor index, Tensor src) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src);
+};
+
+struct TORCH_API scatter__src {
+  using schema = at::Tensor & (at::Tensor &, int64_t, const at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter_";
+  static constexpr const char* overload_name = "src";
+  static constexpr const char* schema_str = "scatter_.src(Tensor(a!) self, int dim, Tensor index, Tensor src) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src);
+};
+
+struct TORCH_API scatter_src_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, const at::Tensor &, const at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter";
+  static constexpr const char* overload_name = "src_out";
+  static constexpr const char* schema_str = "scatter.src_out(Tensor self, int dim, Tensor index, Tensor src, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, at::Tensor & out);
+};
+
+struct TORCH_API scatter_value {
+  using schema = at::Tensor (const at::Tensor &, int64_t, const at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter";
+  static constexpr const char* overload_name = "value";
+  static constexpr const char* schema_str = "scatter.value(Tensor self, int dim, Tensor index, Scalar value) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value);
+};
+
+struct TORCH_API scatter__value {
+  using schema = at::Tensor & (at::Tensor &, int64_t, const at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter_";
+  static constexpr const char* overload_name = "value";
+  static constexpr const char* schema_str = "scatter_.value(Tensor(a!) self, int dim, Tensor index, Scalar value) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value);
+};
+
+struct TORCH_API scatter_value_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, const at::Tensor &, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter";
+  static constexpr const char* overload_name = "value_out";
+  static constexpr const char* schema_str = "scatter.value_out(Tensor self, int dim, Tensor index, Scalar value, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, at::Tensor & out);
+};
+
+struct TORCH_API scatter_reduce {
+  using schema = at::Tensor (const at::Tensor &, int64_t, const at::Tensor &, const at::Tensor &, c10::string_view);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter";
+  static constexpr const char* overload_name = "reduce";
+  static constexpr const char* schema_str = "scatter.reduce(Tensor self, int dim, Tensor index, Tensor src, *, str reduce) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce);
+};
+
+struct TORCH_API scatter__reduce {
+  using schema = at::Tensor & (at::Tensor &, int64_t, const at::Tensor &, const at::Tensor &, c10::string_view);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter_";
+  static constexpr const char* overload_name = "reduce";
+  static constexpr const char* schema_str = "scatter_.reduce(Tensor(a!) self, int dim, Tensor index, Tensor src, *, str reduce) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce);
+};
+
+struct TORCH_API scatter_reduce_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, const at::Tensor &, const at::Tensor &, c10::string_view, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter";
+  static constexpr const char* overload_name = "reduce_out";
+  static constexpr const char* schema_str = "scatter.reduce_out(Tensor self, int dim, Tensor index, Tensor src, *, str reduce, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, at::Tensor & out);
+};
+
+struct TORCH_API scatter_value_reduce {
+  using schema = at::Tensor (const at::Tensor &, int64_t, const at::Tensor &, const at::Scalar &, c10::string_view);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter";
+  static constexpr const char* overload_name = "value_reduce";
+  static constexpr const char* schema_str = "scatter.value_reduce(Tensor self, int dim, Tensor index, Scalar value, *, str reduce) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, c10::string_view reduce);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, c10::string_view reduce);
+};
+
+struct TORCH_API scatter__value_reduce {
+  using schema = at::Tensor & (at::Tensor &, int64_t, const at::Tensor &, const at::Scalar &, c10::string_view);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter_";
+  static constexpr const char* overload_name = "value_reduce";
+  static constexpr const char* schema_str = "scatter_.value_reduce(Tensor(a!) self, int dim, Tensor index, Scalar value, *, str reduce) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, c10::string_view reduce);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, c10::string_view reduce);
+};
+
+struct TORCH_API scatter_value_reduce_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, const at::Tensor &, const at::Scalar &, c10::string_view, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter";
+  static constexpr const char* overload_name = "value_reduce_out";
+  static constexpr const char* schema_str = "scatter.value_reduce_out(Tensor self, int dim, Tensor index, Scalar value, *, str reduce, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, c10::string_view reduce, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Scalar & value, c10::string_view reduce, at::Tensor & out);
+};
+
+struct TORCH_API scatter_dimname_src {
+  using schema = at::Tensor (const at::Tensor &, at::Dimname, const at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter";
+  static constexpr const char* overload_name = "dimname_src";
+  static constexpr const char* schema_str = "scatter.dimname_src(Tensor self, Dimname dim, Tensor index, Tensor src) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::Dimname dim, const at::Tensor & index, const at::Tensor & src);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Dimname dim, const at::Tensor & index, const at::Tensor & src);
+};
+
+struct TORCH_API scatter_dimname_value {
+  using schema = at::Tensor (const at::Tensor &, at::Dimname, const at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::scatter";
+  static constexpr const char* overload_name = "dimname_value";
+  static constexpr const char* schema_str = "scatter.dimname_value(Tensor self, Dimname dim, Tensor index, Scalar value) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::Dimname dim, const at::Tensor & index, const at::Scalar & value);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Dimname dim, const at::Tensor & index, const at::Scalar & value);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_reduce_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_reduce_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..240681f3864c96bc9acaa270e141ef1064b0fce3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_reduce_cpu_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 scatter_reduce(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, bool include_self=true);
+TORCH_API at::Tensor & scatter_reduce_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, bool include_self=true);
+TORCH_API at::Tensor & scatter_reduce_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, bool include_self, at::Tensor & out);
+TORCH_API at::Tensor & scatter_reduce_(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, bool include_self=true);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_reduce_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_reduce_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..60c37ddd2bbd2a18ddbc24b5b35559b880c93ae8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_reduce_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 scatter_reduce(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, bool include_self=true);
+TORCH_API at::Tensor & scatter_reduce_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, bool include_self=true);
+TORCH_API at::Tensor & scatter_reduce_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, bool include_self, at::Tensor & out);
+TORCH_API at::Tensor & scatter_reduce_(at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, bool include_self=true);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_reduce_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_reduce_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..360731c010bb3773e42ae4cac6af990701f89ccb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/scatter_reduce_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/scatter_reduce_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_scatter_reduce_two : public at::meta::structured_scatter_reduce_two {
+void impl(const at::Tensor & self, int64_t dim, const at::Tensor & index, const at::Tensor & src, c10::string_view reduce, bool include_self, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/searchsorted_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/searchsorted_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f32229f1963ede1a007559289bf39df459557648
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/searchsorted_cuda_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 searchsorted(const at::Tensor & sorted_sequence, const at::Tensor & self, bool out_int32=false, bool right=false, ::std::optional<c10::string_view> side=::std::nullopt, const ::std::optional<at::Tensor> & sorter={});
+TORCH_API at::Tensor & searchsorted_out(at::Tensor & out, const at::Tensor & sorted_sequence, const at::Tensor & self, bool out_int32=false, bool right=false, ::std::optional<c10::string_view> side=::std::nullopt, const ::std::optional<at::Tensor> & sorter={});
+TORCH_API at::Tensor & searchsorted_outf(const at::Tensor & sorted_sequence, const at::Tensor & self, bool out_int32, bool right, ::std::optional<c10::string_view> side, const ::std::optional<at::Tensor> & sorter, at::Tensor & out);
+TORCH_API at::Tensor searchsorted(const at::Tensor & sorted_sequence, const at::Scalar & self, bool out_int32=false, bool right=false, ::std::optional<c10::string_view> side=::std::nullopt, const ::std::optional<at::Tensor> & sorter={});
+TORCH_API at::Tensor & searchsorted_out(at::Tensor & out, const at::Tensor & sorted_sequence, const at::Scalar & self, bool out_int32=false, bool right=false, ::std::optional<c10::string_view> side=::std::nullopt, const ::std::optional<at::Tensor> & sorter={});
+TORCH_API at::Tensor & searchsorted_outf(const at::Tensor & sorted_sequence, const at::Scalar & self, bool out_int32, bool right, ::std::optional<c10::string_view> side, const ::std::optional<at::Tensor> & sorter, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/select_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/select_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d8651042603ddebc02a2486e90f443c92634c14c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/select_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 select(const at::Tensor & self, int64_t dim, int64_t index);
+TORCH_API at::Tensor select_symint(const at::Tensor & self, int64_t dim, c10::SymInt index);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/select_copy_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/select_copy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..db735718c52dff07f65388f8c8a65afb9f83da66
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/select_copy_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & select_copy_int_out_symint(const at::Tensor & self, int64_t dim, c10::SymInt index, at::Tensor & out);
+TORCH_API at::Tensor select_copy_sparse_csr(const at::Tensor & self, int64_t dim, int64_t index);
+TORCH_API at::Tensor select_copy_symint(const at::Tensor & self, int64_t dim, c10::SymInt index);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/select_scatter.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/select_scatter.h
new file mode 100644
index 0000000000000000000000000000000000000000..094568ec90862e26e9962c04e664ffa72d988362
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/select_scatter.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/select_scatter_ops.h>
+
+namespace at {
+
+
+// aten::select_scatter(Tensor self, Tensor src, int dim, SymInt index) -> Tensor
+inline at::Tensor select_scatter(const at::Tensor & self, const at::Tensor & src, int64_t dim, int64_t index) {
+    return at::_ops::select_scatter::call(self, src, dim, index);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor select_scatter(const at::Tensor & self, const at::Tensor & src, int64_t dim, int64_t index) {
+    return at::_ops::select_scatter::call(self, src, dim, index);
+  }
+}
+
+// aten::select_scatter(Tensor self, Tensor src, int dim, SymInt index) -> Tensor
+inline at::Tensor select_scatter_symint(const at::Tensor & self, const at::Tensor & src, int64_t dim, c10::SymInt index) {
+    return at::_ops::select_scatter::call(self, src, dim, index);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor select_scatter(const at::Tensor & self, const at::Tensor & src, int64_t dim, c10::SymInt index) {
+    return at::_ops::select_scatter::call(self, src, dim, index);
+  }
+}
+
+// aten::select_scatter.out(Tensor self, Tensor src, int dim, SymInt index, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & select_scatter_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & src, int64_t dim, int64_t index) {
+    return at::_ops::select_scatter_out::call(self, src, dim, index, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & select_scatter_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & src, int64_t dim, int64_t index) {
+    return at::_ops::select_scatter_out::call(self, src, dim, index, out);
+  }
+}
+
+// aten::select_scatter.out(Tensor self, Tensor src, int dim, SymInt index, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & select_scatter_outf(const at::Tensor & self, const at::Tensor & src, int64_t dim, int64_t index, at::Tensor & out) {
+    return at::_ops::select_scatter_out::call(self, src, dim, index, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & select_scatter_outf(const at::Tensor & self, const at::Tensor & src, int64_t dim, int64_t index, at::Tensor & out) {
+    return at::_ops::select_scatter_out::call(self, src, dim, index, out);
+  }
+}
+
+// aten::select_scatter.out(Tensor self, Tensor src, int dim, SymInt index, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & select_scatter_symint_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & src, int64_t dim, c10::SymInt index) {
+    return at::_ops::select_scatter_out::call(self, src, dim, index, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & select_scatter_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & src, int64_t dim, c10::SymInt index) {
+    return at::_ops::select_scatter_out::call(self, src, dim, index, out);
+  }
+}
+
+// aten::select_scatter.out(Tensor self, Tensor src, int dim, SymInt index, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & select_scatter_symint_outf(const at::Tensor & self, const at::Tensor & src, int64_t dim, c10::SymInt index, at::Tensor & out) {
+    return at::_ops::select_scatter_out::call(self, src, dim, index, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & select_scatter_outf(const at::Tensor & self, const at::Tensor & src, int64_t dim, c10::SymInt index, at::Tensor & out) {
+    return at::_ops::select_scatter_out::call(self, src, dim, index, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/selu.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/selu.h
new file mode 100644
index 0000000000000000000000000000000000000000..2d53030059b3234815c5eaa9a6287773d7b3b955
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/selu.h
@@ -0,0 +1,41 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/selu_ops.h>
+
+namespace at {
+
+
+// aten::selu(Tensor self) -> Tensor
+inline at::Tensor selu(const at::Tensor & self) {
+    return at::_ops::selu::call(self);
+}
+
+// aten::selu_(Tensor(a!) self) -> Tensor(a!)
+inline at::Tensor & selu_(at::Tensor & self) {
+    return at::_ops::selu_::call(self);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/set_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/set_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..49701c440080348ec412c01b51dcd31a4b1c7744
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/set_ops.h
@@ -0,0 +1,166 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 set__source_Storage {
+  using schema = at::Tensor & (at::Tensor &, at::Storage);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::set_";
+  static constexpr const char* overload_name = "source_Storage";
+  static constexpr const char* schema_str = "set_.source_Storage(Tensor(a!) self, Storage source) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, at::Storage source);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, at::Storage source);
+};
+
+struct TORCH_API set__source_Storage_storage_offset {
+  using schema = at::Tensor & (at::Tensor &, at::Storage, c10::SymInt, c10::SymIntArrayRef, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::set_";
+  static constexpr const char* overload_name = "source_Storage_storage_offset";
+  static constexpr const char* schema_str = "set_.source_Storage_storage_offset(Tensor(a!) self, Storage source, SymInt storage_offset, SymInt[] size, SymInt[] stride=[]) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, at::Storage source, c10::SymInt storage_offset, c10::SymIntArrayRef size, c10::SymIntArrayRef stride);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, at::Storage source, c10::SymInt storage_offset, c10::SymIntArrayRef size, c10::SymIntArrayRef stride);
+};
+
+struct TORCH_API set__source_Tensor_storage_offset {
+  using schema = at::Tensor & (at::Tensor &, const at::Tensor &, c10::SymInt, c10::SymIntArrayRef, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::set_";
+  static constexpr const char* overload_name = "source_Tensor_storage_offset";
+  static constexpr const char* schema_str = "set_.source_Tensor_storage_offset(Tensor(a!) self, Tensor source, SymInt storage_offset, SymInt[] size, SymInt[] stride=[]) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Tensor & source, c10::SymInt storage_offset, c10::SymIntArrayRef size, c10::SymIntArrayRef stride);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Tensor & source, c10::SymInt storage_offset, c10::SymIntArrayRef size, c10::SymIntArrayRef stride);
+};
+
+struct TORCH_API set__source_Tensor {
+  using schema = at::Tensor & (at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::set_";
+  static constexpr const char* overload_name = "source_Tensor";
+  static constexpr const char* schema_str = "set_.source_Tensor(Tensor(a!) self, Tensor source) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Tensor & source);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Tensor & source);
+};
+
+struct TORCH_API set_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::set_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "set_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API set_source_Storage_out {
+  using schema = at::Tensor & (const at::Tensor &, at::Storage, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::set";
+  static constexpr const char* overload_name = "source_Storage_out";
+  static constexpr const char* schema_str = "set.source_Storage_out(Tensor self, Storage source, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, at::Storage source, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Storage source, at::Tensor & out);
+};
+
+struct TORCH_API set_source_Storage {
+  using schema = at::Tensor (const at::Tensor &, at::Storage);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::set";
+  static constexpr const char* overload_name = "source_Storage";
+  static constexpr const char* schema_str = "set.source_Storage(Tensor self, Storage source) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::Storage source);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Storage source);
+};
+
+struct TORCH_API set_source_Storage_storage_offset_out {
+  using schema = at::Tensor & (const at::Tensor &, at::Storage, c10::SymInt, c10::SymIntArrayRef, c10::SymIntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::set";
+  static constexpr const char* overload_name = "source_Storage_storage_offset_out";
+  static constexpr const char* schema_str = "set.source_Storage_storage_offset_out(Tensor self, Storage source, SymInt storage_offset, SymInt[] size, SymInt[] stride=[], *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, at::Storage source, c10::SymInt storage_offset, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Storage source, c10::SymInt storage_offset, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, at::Tensor & out);
+};
+
+struct TORCH_API set_source_Storage_storage_offset {
+  using schema = at::Tensor (const at::Tensor &, at::Storage, c10::SymInt, c10::SymIntArrayRef, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::set";
+  static constexpr const char* overload_name = "source_Storage_storage_offset";
+  static constexpr const char* schema_str = "set.source_Storage_storage_offset(Tensor self, Storage source, SymInt storage_offset, SymInt[] size, SymInt[] stride=[]) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, at::Storage source, c10::SymInt storage_offset, c10::SymIntArrayRef size, c10::SymIntArrayRef stride);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Storage source, c10::SymInt storage_offset, c10::SymIntArrayRef size, c10::SymIntArrayRef stride);
+};
+
+struct TORCH_API set_source_Tensor_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 const char* name = "aten::set";
+  static constexpr const char* overload_name = "source_Tensor_out";
+  static constexpr const char* schema_str = "set.source_Tensor_out(Tensor self, Tensor source, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & source, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & source, at::Tensor & out);
+};
+
+struct TORCH_API set_source_Tensor {
+  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 const char* name = "aten::set";
+  static constexpr const char* overload_name = "source_Tensor";
+  static constexpr const char* schema_str = "set.source_Tensor(Tensor self, Tensor source) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & source);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & source);
+};
+
+struct TORCH_API set_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 const char* name = "aten::set";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "set.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);
+};
+
+struct TORCH_API set {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::set";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "set(Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sgn_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sgn_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..5ddbadac45bd50ec4e927b9544c174b33ff5fde4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sgn_native.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/sgn_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_sgn_out : public at::meta::structured_sgn {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+TORCH_API at::Tensor NestedTensor_sgn(const at::Tensor & self);
+TORCH_API at::Tensor & NestedTensor_sgn_(at::Tensor & self);
+TORCH_API at::Tensor sgn_sparse(const at::Tensor & self);
+TORCH_API at::Tensor & sgn_sparse_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & sgn_sparse_(at::Tensor & self);
+TORCH_API at::Tensor sgn_sparse_csr(const at::Tensor & self);
+TORCH_API at::Tensor & sgn_sparse_csr_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & sgn_sparse_csr_(at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..74f4e4aa25f333e542c60f160eb6a765dad787e8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_backward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 sigmoid_backward_grad_input {
+  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 const char* name = "aten::sigmoid_backward";
+  static constexpr const char* overload_name = "grad_input";
+  static constexpr const char* schema_str = "sigmoid_backward.grad_input(Tensor grad_output, Tensor output, *, Tensor(a!) grad_input) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & grad_output, const at::Tensor & output, at::Tensor & grad_input);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & output, at::Tensor & grad_input);
+};
+
+struct TORCH_API sigmoid_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 const char* name = "aten::sigmoid_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "sigmoid_backward(Tensor grad_output, Tensor output) -> Tensor";
+  static at::Tensor call(const at::Tensor & grad_output, const at::Tensor & output);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & output);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b346b0e6f98af8d52b88ede7b9f0f348a9ee55bf
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sigmoid(const at::Tensor & self);
+TORCH_API at::Tensor & sigmoid_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..805a812d676f2d8f3b0d349bc9c2b11f54ca5a05
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sigmoid(const at::Tensor & self);
+TORCH_API at::Tensor & sigmoid_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & sigmoid_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & sigmoid_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..c5f26ce4d7ced81f9dea0c782ef8667252df8d13
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_native.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/sigmoid_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_sigmoid_out : public at::meta::structured_sigmoid {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+TORCH_API at::Tensor mkldnn_sigmoid(const at::Tensor & self);
+TORCH_API at::Tensor & mkldnn_sigmoid_(at::Tensor & self);
+TORCH_API at::Tensor sigmoid_quantized_cpu(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..4cffec1458c4d6b03f56f6d53b3202acae151ee0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sigmoid_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 sigmoid {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sigmoid";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "sigmoid(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 sigmoid_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sigmoid_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "sigmoid_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API sigmoid_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 const char* name = "aten::sigmoid";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "sigmoid.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sign_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sign_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..af9936e1059bba2de8ec47b37ffa2fa2c7d2706c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sign_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sign(const at::Tensor & self);
+TORCH_API at::Tensor & sign_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & sign_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & sign_(at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sin_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sin_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c36bdd14636c3cfc32dfcd9d9d4a0eed830db6ee
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sin_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sin(const at::Tensor & self);
+TORCH_API at::Tensor & sin_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sin_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sin_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..059f119f203036e24dfbbd2d528e33295747108a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sin_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sin(const at::Tensor & self);
+TORCH_API at::Tensor & sin_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & sin_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & sin_(at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sin_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sin_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..30af8448d816827384cca76447249b4b3c3b34a2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sin_native.h
@@ -0,0 +1,35 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/sin_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_sin_out : public at::meta::structured_sin {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+TORCH_API at::Tensor NestedTensor_sin(const at::Tensor & self);
+TORCH_API at::Tensor sin_sparse(const at::Tensor & self);
+TORCH_API at::Tensor & sin_sparse_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & sin_sparse_(at::Tensor & self);
+TORCH_API at::Tensor sin_sparse_csr(const at::Tensor & self);
+TORCH_API at::Tensor & sin_sparse_csr_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & sin_sparse_csr_(at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinc_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinc_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..818dcfdb4b46c387d1f389b544308a518dbf269f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinc_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sinc(const at::Tensor & self);
+TORCH_API at::Tensor & sinc_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinc_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinc_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..de99368d652ec3bb48cce605993ff8c5ae4f5d73
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinc_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sinc(const at::Tensor & self);
+TORCH_API at::Tensor & sinc_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & sinc_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & sinc_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinc_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinc_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b4c2a9cfe4ea6377183577c588b457db347a7586
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinc_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/sinc_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_sinc_out : public at::meta::structured_sinc {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinh_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinh_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..71ce2a05fe2d5ea3aaf1f5e0097c797c56b49139
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sinh_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sinh(const at::Tensor & self);
+TORCH_API at::Tensor & sinh_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & sinh_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & sinh_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/size_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/size_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..79827da20501da1fc2eedcebd880cd82e59bf4aa
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/size_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 size_int {
+  using schema = int64_t (const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::size";
+  static constexpr const char* overload_name = "int";
+  static constexpr const char* schema_str = "size.int(Tensor self, int dim) -> int";
+  static int64_t call(const at::Tensor & self, int64_t dim);
+  static int64_t redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim);
+};
+
+struct TORCH_API size_Dimname {
+  using schema = int64_t (const at::Tensor &, at::Dimname);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::size";
+  static constexpr const char* overload_name = "Dimname";
+  static constexpr const char* schema_str = "size.Dimname(Tensor self, Dimname dim) -> int";
+  static int64_t call(const at::Tensor & self, at::Dimname dim);
+  static int64_t redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Dimname dim);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_copy_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_copy_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..a002b127d1db604e6f78abbf3f89a763cdf1b8f3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_copy_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 slice_copy_Tensor {
+  using schema = at::Tensor (const at::Tensor &, int64_t, ::std::optional<c10::SymInt>, ::std::optional<c10::SymInt>, c10::SymInt);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::slice_copy";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "slice_copy.Tensor(Tensor self, int dim=0, SymInt? start=None, SymInt? end=None, SymInt step=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t dim, ::std::optional<c10::SymInt> start, ::std::optional<c10::SymInt> end, c10::SymInt step);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, ::std::optional<c10::SymInt> start, ::std::optional<c10::SymInt> end, c10::SymInt step);
+};
+
+struct TORCH_API slice_copy_Tensor_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, ::std::optional<c10::SymInt>, ::std::optional<c10::SymInt>, c10::SymInt, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::slice_copy";
+  static constexpr const char* overload_name = "Tensor_out";
+  static constexpr const char* schema_str = "slice_copy.Tensor_out(Tensor self, int dim=0, SymInt? start=None, SymInt? end=None, SymInt step=1, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, int64_t dim, ::std::optional<c10::SymInt> start, ::std::optional<c10::SymInt> end, c10::SymInt step, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, ::std::optional<c10::SymInt> start, ::std::optional<c10::SymInt> end, c10::SymInt step, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_inverse_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_inverse_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..653767db3160e7c1c23bed481151a298d84f72d7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_inverse_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 slice_inverse {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, int64_t, ::std::optional<c10::SymInt>, ::std::optional<c10::SymInt>, c10::SymInt);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::slice_inverse";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "slice_inverse(Tensor(a) self, Tensor src, int dim=0, SymInt? start=None, SymInt? end=None, SymInt step=1) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & src, int64_t dim, ::std::optional<c10::SymInt> start, ::std::optional<c10::SymInt> end, c10::SymInt step);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & src, int64_t dim, ::std::optional<c10::SymInt> start, ::std::optional<c10::SymInt> end, c10::SymInt step);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..aa85c91ca1a3a05160c27bfbb3ea7634f5f4190a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 slice_Tensor {
+  using schema = at::Tensor (const at::Tensor &, int64_t, ::std::optional<c10::SymInt>, ::std::optional<c10::SymInt>, c10::SymInt);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::slice";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "slice.Tensor(Tensor(a) self, int dim=0, SymInt? start=None, SymInt? end=None, SymInt step=1) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, int64_t dim, ::std::optional<c10::SymInt> start, ::std::optional<c10::SymInt> end, c10::SymInt step);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, ::std::optional<c10::SymInt> start, ::std::optional<c10::SymInt> end, c10::SymInt step);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_scatter_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_scatter_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b186bfa3d4d7806e8108ad2285be9c9d38765595
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slice_scatter_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & slice_scatter_out_symint(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);
+TORCH_API at::Tensor slice_scatter(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);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slow_conv3d_forward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slow_conv3d_forward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..81780890c21ed93a5a76fc5a448f00adbf6580a8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slow_conv3d_forward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 slow_conv3d_forward_output {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::slow_conv3d_forward";
+  static constexpr const char* overload_name = "output";
+  static constexpr const char* schema_str = "slow_conv3d_forward.output(Tensor self, Tensor weight, SymInt[3] kernel_size, Tensor? bias, SymInt[3] stride, SymInt[3] padding, *, Tensor(a!) output) -> Tensor(a!)";
+  static at::Tensor & call(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);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, 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);
+};
+
+struct TORCH_API slow_conv3d_forward {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::slow_conv3d_forward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "slow_conv3d_forward(Tensor self, Tensor weight, SymInt[3] kernel_size, Tensor? bias, SymInt[3] stride, SymInt[3] padding) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slow_conv_dilated2d.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slow_conv_dilated2d.h
new file mode 100644
index 0000000000000000000000000000000000000000..eb9283bdf174936484d1dce2ff2e5989d5a7743f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slow_conv_dilated2d.h
@@ -0,0 +1,97 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/slow_conv_dilated2d_ops.h>
+
+namespace at {
+
+
+// aten::slow_conv_dilated2d(Tensor self, Tensor weight, SymInt[2] kernel_size, Tensor? bias=None, SymInt[2] stride=1, SymInt[2] padding=0, SymInt[2] dilation=1) -> Tensor
+inline at::Tensor slow_conv_dilated2d(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias={}, at::IntArrayRef stride=1, at::IntArrayRef padding=0, at::IntArrayRef dilation=1) {
+    return at::_ops::slow_conv_dilated2d::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation));
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor slow_conv_dilated2d(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias={}, at::IntArrayRef stride=1, at::IntArrayRef padding=0, at::IntArrayRef dilation=1) {
+    return at::_ops::slow_conv_dilated2d::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation));
+  }
+}
+
+// aten::slow_conv_dilated2d(Tensor self, Tensor weight, SymInt[2] kernel_size, Tensor? bias=None, SymInt[2] stride=1, SymInt[2] padding=0, SymInt[2] dilation=1) -> Tensor
+inline at::Tensor slow_conv_dilated2d_symint(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias={}, c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef dilation=c10::SymInt(1)) {
+    return at::_ops::slow_conv_dilated2d::call(self, weight, kernel_size, bias, stride, padding, dilation);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor slow_conv_dilated2d(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias={}, c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef dilation=c10::SymInt(1)) {
+    return at::_ops::slow_conv_dilated2d::call(self, weight, kernel_size, bias, stride, padding, dilation);
+  }
+}
+
+// aten::slow_conv_dilated2d.out(Tensor self, Tensor weight, SymInt[2] kernel_size, Tensor? bias=None, SymInt[2] stride=1, SymInt[2] padding=0, SymInt[2] dilation=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & slow_conv_dilated2d_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias={}, at::IntArrayRef stride=1, at::IntArrayRef padding=0, at::IntArrayRef dilation=1) {
+    return at::_ops::slow_conv_dilated2d_out::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & slow_conv_dilated2d_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias={}, at::IntArrayRef stride=1, at::IntArrayRef padding=0, at::IntArrayRef dilation=1) {
+    return at::_ops::slow_conv_dilated2d_out::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation), out);
+  }
+}
+
+// aten::slow_conv_dilated2d.out(Tensor self, Tensor weight, SymInt[2] kernel_size, Tensor? bias=None, SymInt[2] stride=1, SymInt[2] padding=0, SymInt[2] dilation=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & slow_conv_dilated2d_outf(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, at::Tensor & out) {
+    return at::_ops::slow_conv_dilated2d_out::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & slow_conv_dilated2d_outf(const at::Tensor & self, const at::Tensor & weight, at::IntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, at::Tensor & out) {
+    return at::_ops::slow_conv_dilated2d_out::call(self, weight, c10::fromIntArrayRefSlow(kernel_size), bias, c10::fromIntArrayRefSlow(stride), c10::fromIntArrayRefSlow(padding), c10::fromIntArrayRefSlow(dilation), out);
+  }
+}
+
+// aten::slow_conv_dilated2d.out(Tensor self, Tensor weight, SymInt[2] kernel_size, Tensor? bias=None, SymInt[2] stride=1, SymInt[2] padding=0, SymInt[2] dilation=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & slow_conv_dilated2d_symint_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias={}, c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef dilation=c10::SymInt(1)) {
+    return at::_ops::slow_conv_dilated2d_out::call(self, weight, kernel_size, bias, stride, padding, dilation, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & slow_conv_dilated2d_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias={}, c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef dilation=c10::SymInt(1)) {
+    return at::_ops::slow_conv_dilated2d_out::call(self, weight, kernel_size, bias, stride, padding, dilation, out);
+  }
+}
+
+// aten::slow_conv_dilated2d.out(Tensor self, Tensor weight, SymInt[2] kernel_size, Tensor? bias=None, SymInt[2] stride=1, SymInt[2] padding=0, SymInt[2] dilation=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & slow_conv_dilated2d_symint_outf(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, at::Tensor & out) {
+    return at::_ops::slow_conv_dilated2d_out::call(self, weight, kernel_size, bias, stride, padding, dilation, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & slow_conv_dilated2d_outf(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, at::Tensor & out) {
+    return at::_ops::slow_conv_dilated2d_out::call(self, weight, kernel_size, bias, stride, padding, dilation, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slow_conv_transpose3d_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slow_conv_transpose3d_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..e9a221be31fb28a4c929bd72e36a7749ace4c483
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/slow_conv_transpose3d_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 slow_conv_transpose3d_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::slow_conv_transpose3d";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "slow_conv_transpose3d.out(Tensor self, Tensor weight, SymInt[3] kernel_size, Tensor? bias=None, SymInt[3] stride=1, SymInt[3] padding=0, SymInt[3] output_padding=0, SymInt[3] dilation=1, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef dilation, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef dilation, at::Tensor & out);
+};
+
+struct TORCH_API slow_conv_transpose3d {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::slow_conv_transpose3d";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "slow_conv_transpose3d(Tensor self, Tensor weight, SymInt[3] kernel_size, Tensor? bias=None, SymInt[3] stride=1, SymInt[3] padding=0, SymInt[3] output_padding=0, SymInt[3] dilation=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef dilation);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const ::std::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef output_padding, c10::SymIntArrayRef dilation);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/smooth_l1_loss_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/smooth_l1_loss_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8bf02eb52801ac8baba338bc53ddbd0e576ead17
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/smooth_l1_loss_backward_cpu_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & smooth_l1_loss_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & target, int64_t reduction, double beta);
+TORCH_API at::Tensor & smooth_l1_loss_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & target, int64_t reduction, double beta, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/smooth_l1_loss_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/smooth_l1_loss_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..d94b24d6bafb66fae711edd4da4bad2acbf9ec8b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/smooth_l1_loss_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 smooth_l1_loss_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, int64_t, double, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::smooth_l1_loss";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "smooth_l1_loss.out(Tensor self, Tensor target, int reduction=Mean, float beta=1.0, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & target, int64_t reduction, double beta, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & target, int64_t reduction, double beta, at::Tensor & out);
+};
+
+struct TORCH_API smooth_l1_loss {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, int64_t, double);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::smooth_l1_loss";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "smooth_l1_loss(Tensor self, Tensor target, int reduction=Mean, float beta=1.0) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & target, int64_t reduction, double beta);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & target, int64_t reduction, double beta);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softplus.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softplus.h
new file mode 100644
index 0000000000000000000000000000000000000000..b30de5a31dc1c5f623d4b2e5016ccb8d237d0b11
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softplus.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/softplus_ops.h>
+
+namespace at {
+
+
+// aten::softplus.out(Tensor self, Scalar beta=1, Scalar threshold=20, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & softplus_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & beta=1, const at::Scalar & threshold=20) {
+    return at::_ops::softplus_out::call(self, beta, threshold, out);
+}
+// aten::softplus.out(Tensor self, Scalar beta=1, Scalar threshold=20, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & softplus_outf(const at::Tensor & self, const at::Scalar & beta, const at::Scalar & threshold, at::Tensor & out) {
+    return at::_ops::softplus_out::call(self, beta, threshold, out);
+}
+
+// aten::softplus(Tensor self, Scalar beta=1, Scalar threshold=20) -> Tensor
+inline at::Tensor softplus(const at::Tensor & self, const at::Scalar & beta=1, const at::Scalar & threshold=20) {
+    return at::_ops::softplus::call(self, beta, threshold);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softplus_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softplus_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5d55aadc9cd745fe44c7365c7f225e8acdc06d9e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softplus_backward_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 softplus_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & beta, const at::Scalar & threshold);
+TORCH_API at::Tensor & softplus_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & beta, const at::Scalar & threshold);
+TORCH_API at::Tensor & softplus_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & beta, const at::Scalar & threshold, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softshrink_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softshrink_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3e0efb7642988d25e209b86ed1ac47a85d8416d4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softshrink_backward_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 softshrink_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & lambd);
+TORCH_API at::Tensor & softshrink_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & lambd);
+TORCH_API at::Tensor & softshrink_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & lambd, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softshrink_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softshrink_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..d05ccfce6f26d8e5782067c8ace4a6aad06c7c15
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softshrink_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_softshrink : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & self, const at::Scalar & lambd);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softshrink_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softshrink_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..5232709bed73287c3d403e3ce24d0350f3cdaa74
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/softshrink_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/softshrink_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_softshrink_out : public at::meta::structured_softshrink {
+void impl(const at::Tensor & self, const at::Scalar & lambd, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sort_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sort_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..0bc3dea17b87652d1a33b7b08f21004e7f088594
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sort_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,at::Tensor> sort(const at::Tensor & self, at::Dimname dim, bool descending=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> sort_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, at::Dimname dim, bool descending=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> sort_outf(const at::Tensor & self, at::Dimname dim, bool descending, at::Tensor & values, at::Tensor & indices);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> sort(const at::Tensor & self, ::std::optional<bool> stable, at::Dimname dim, bool descending=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> sort_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, ::std::optional<bool> stable, at::Dimname dim, bool descending=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> sort_outf(const at::Tensor & self, ::std::optional<bool> stable, at::Dimname dim, bool descending, at::Tensor & values, at::Tensor & indices);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sort_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sort_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..7db065e2087343b78e6aab64ed8a25667df997b7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sort_ops.h
@@ -0,0 +1,111 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 sort_values {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, int64_t, bool, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sort";
+  static constexpr const char* overload_name = "values";
+  static constexpr const char* schema_str = "sort.values(Tensor self, int dim=-1, bool descending=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, int64_t dim, bool descending, at::Tensor & values, at::Tensor & indices);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, bool descending, at::Tensor & values, at::Tensor & indices);
+};
+
+struct TORCH_API sort_values_stable {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, ::std::optional<bool>, int64_t, bool, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sort";
+  static constexpr const char* overload_name = "values_stable";
+  static constexpr const char* schema_str = "sort.values_stable(Tensor self, *, bool? stable, int dim=-1, bool descending=False, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, ::std::optional<bool> stable, int64_t dim, bool descending, at::Tensor & values, at::Tensor & indices);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<bool> stable, int64_t dim, bool descending, at::Tensor & values, at::Tensor & indices);
+};
+
+struct TORCH_API sort {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, int64_t, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sort";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "sort(Tensor self, int dim=-1, bool descending=False) -> (Tensor values, Tensor indices)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, int64_t dim, bool descending);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, bool descending);
+};
+
+struct TORCH_API sort_stable {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, ::std::optional<bool>, int64_t, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sort";
+  static constexpr const char* overload_name = "stable";
+  static constexpr const char* schema_str = "sort.stable(Tensor self, *, bool? stable, int dim=-1, bool descending=False) -> (Tensor values, Tensor indices)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, ::std::optional<bool> stable, int64_t dim, bool descending);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<bool> stable, int64_t dim, bool descending);
+};
+
+struct TORCH_API sort_dimname_values {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, at::Dimname, bool, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sort";
+  static constexpr const char* overload_name = "dimname_values";
+  static constexpr const char* schema_str = "sort.dimname_values(Tensor self, Dimname dim, bool descending=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, at::Dimname dim, bool descending, at::Tensor & values, at::Tensor & indices);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Dimname dim, bool descending, at::Tensor & values, at::Tensor & indices);
+};
+
+struct TORCH_API sort_dimname_values_stable {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, ::std::optional<bool>, at::Dimname, bool, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sort";
+  static constexpr const char* overload_name = "dimname_values_stable";
+  static constexpr const char* schema_str = "sort.dimname_values_stable(Tensor self, *, bool? stable, Dimname dim, bool descending=False, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices)";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, ::std::optional<bool> stable, at::Dimname dim, bool descending, at::Tensor & values, at::Tensor & indices);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<bool> stable, at::Dimname dim, bool descending, at::Tensor & values, at::Tensor & indices);
+};
+
+struct TORCH_API sort_dimname {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, at::Dimname, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sort";
+  static constexpr const char* overload_name = "dimname";
+  static constexpr const char* schema_str = "sort.dimname(Tensor self, Dimname dim, bool descending=False) -> (Tensor values, Tensor indices)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, at::Dimname dim, bool descending);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Dimname dim, bool descending);
+};
+
+struct TORCH_API sort_dimname_stable {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, ::std::optional<bool>, at::Dimname, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sort";
+  static constexpr const char* overload_name = "dimname_stable";
+  static constexpr const char* schema_str = "sort.dimname_stable(Tensor self, *, bool? stable, Dimname dim, bool descending=False) -> (Tensor values, Tensor indices)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, ::std::optional<bool> stable, at::Dimname dim, bool descending);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<bool> stable, at::Dimname dim, bool descending);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_bsr_tensor_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_bsr_tensor_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4563f54a756e81aec001d7f1bb65e01e08f97c81
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_bsr_tensor_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor sparse_bsr_tensor(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, at::TensorOptions options);
+TORCH_API at::Tensor sparse_bsr_tensor(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+TORCH_API at::Tensor sparse_bsr_tensor(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::TensorOptions options);
+TORCH_API at::Tensor sparse_bsr_tensor(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_compressed_tensor_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_compressed_tensor_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..791741f273a379130c25d937e8ff8957cb4461aa
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_compressed_tensor_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sparse_compressed_tensor(const at::Tensor & compressed_indices, const at::Tensor & plain_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor sparse_compressed_tensor(const at::Tensor & compressed_indices, const at::Tensor & plain_indices, const at::Tensor & values, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_coo_tensor_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_coo_tensor_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8e6585cc401f58a147dbed89f2239b165028994b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_coo_tensor_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor sparse_coo_tensor(const at::Tensor & indices, const at::Tensor & values, at::TensorOptions options={}, ::std::optional<bool> is_coalesced=::std::nullopt);
+TORCH_API at::Tensor sparse_coo_tensor(const at::Tensor & indices, const at::Tensor & values, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<bool> is_coalesced);
+TORCH_API at::Tensor sparse_coo_tensor(const at::Tensor & indices, const at::Tensor & values, at::IntArrayRef size, at::TensorOptions options={}, ::std::optional<bool> is_coalesced=::std::nullopt);
+TORCH_API at::Tensor sparse_coo_tensor(const at::Tensor & indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory, ::std::optional<bool> is_coalesced);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_csr_tensor.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_csr_tensor.h
new file mode 100644
index 0000000000000000000000000000000000000000..227ce92d02cee7d825c6e174834f0adcc56c93ec
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_csr_tensor.h
@@ -0,0 +1,49 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/sparse_csr_tensor_ops.h>
+
+namespace at {
+
+
+// aten::sparse_csr_tensor.crow_col_value_size(Tensor crow_indices, Tensor col_indices, Tensor values, int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=False) -> Tensor
+inline at::Tensor sparse_csr_tensor(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, at::TensorOptions options) {
+    return at::_ops::sparse_csr_tensor_crow_col_value_size::call(crow_indices, col_indices, values, size, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+// aten::sparse_csr_tensor.crow_col_value_size(Tensor crow_indices, Tensor col_indices, Tensor values, int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=False) -> Tensor
+inline at::Tensor sparse_csr_tensor(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::sparse_csr_tensor_crow_col_value_size::call(crow_indices, col_indices, values, size, dtype, layout, device, pin_memory);
+}
+
+// aten::sparse_csr_tensor.crow_col_value(Tensor crow_indices, Tensor col_indices, Tensor values, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=False) -> Tensor
+inline at::Tensor sparse_csr_tensor(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::TensorOptions options) {
+    return at::_ops::sparse_csr_tensor_crow_col_value::call(crow_indices, col_indices, values, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+// aten::sparse_csr_tensor.crow_col_value(Tensor crow_indices, Tensor col_indices, Tensor values, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=False) -> Tensor
+inline at::Tensor sparse_csr_tensor(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory) {
+    return at::_ops::sparse_csr_tensor_crow_col_value::call(crow_indices, col_indices, values, dtype, layout, device, pin_memory);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_csr_tensor_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_csr_tensor_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..ce14bd20ff893a2d5d60dd75a6c24da6b7b45bae
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_csr_tensor_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API sparse_csr_tensor_crow_col_value_size {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, at::IntArrayRef, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sparse_csr_tensor";
+  static constexpr const char* overload_name = "crow_col_value_size";
+  static constexpr const char* schema_str = "sparse_csr_tensor.crow_col_value_size(Tensor crow_indices, Tensor col_indices, Tensor values, int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=False) -> Tensor";
+  static at::Tensor call(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+struct TORCH_API sparse_csr_tensor_crow_col_value {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sparse_csr_tensor";
+  static constexpr const char* overload_name = "crow_col_value";
+  static constexpr const char* schema_str = "sparse_csr_tensor.crow_col_value(Tensor crow_indices, Tensor col_indices, Tensor values, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=False) -> Tensor";
+  static at::Tensor call(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_dim_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_dim_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..f914dc872b2d74294fb7ba3b2e9a048f2f4a90b0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sparse_dim_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API sparse_dim {
+  using schema = int64_t (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sparse_dim";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "sparse_dim(Tensor self) -> int";
+  static int64_t call(const at::Tensor & self);
+  static int64_t redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_airy_ai_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_airy_ai_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..555afabd0512a0388f4262153916207dce72bcd4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_airy_ai_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_airy_ai(const at::Tensor & x);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_airy_ai_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_airy_ai_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c09e8c39d1ad8096bc796ae31674a19e9ef4d9f2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_airy_ai_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_airy_ai(const at::Tensor & x);
+TORCH_API at::Tensor & special_airy_ai_out(at::Tensor & out, const at::Tensor & x);
+TORCH_API at::Tensor & special_airy_ai_outf(const at::Tensor & x, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_airy_ai_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_airy_ai_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..1eb8ef38d31289ee946c78f9bc26471aeef7634f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_airy_ai_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_special_airy_ai : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & x);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_j0_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_j0_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..bd227e18fd8411a52977bc99c323e008043abe76
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_j0_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_bessel_j0 {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_bessel_j0";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "special_bessel_j0(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 special_bessel_j0_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 const char* name = "aten::special_bessel_j0";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "special_bessel_j0.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_j1_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_j1_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..cf93ce181b06faa1ce61b94436ef3c69b3aedd27
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_j1_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_bessel_j1(const at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_j1_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_j1_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ba34562c63782f10c7dbe6927fe3a9bc14d96ec9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_j1_cuda_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_bessel_j1(const at::Tensor & self);
+TORCH_API at::Tensor & special_bessel_j1_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & special_bessel_j1_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_y1_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_y1_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..1bc66a6ebc2ab6f913fd8c9032857ba1238418cc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_y1_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_bessel_y1(const at::Tensor & self);
+TORCH_API at::Tensor & special_bessel_y1_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & special_bessel_y1_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_y1_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_y1_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..db572088b7fa5f1f2754e968d6045713f96ef33c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_y1_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_bessel_y1(const at::Tensor & self);
+TORCH_API at::Tensor & special_bessel_y1_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & special_bessel_y1_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_y1_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_y1_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..6ae1bd2ff006f91e5a11bc83191a8edb5e336bf6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_bessel_y1_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_bessel_y1_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_special_bessel_y1_out : public at::meta::structured_special_bessel_y1 {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_t_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_t_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..ff2b273ae346a1122b3fea467970a8e2c7606a0b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_t_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_special_chebyshev_polynomial_t : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & x, const at::Tensor & n);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u.h
new file mode 100644
index 0000000000000000000000000000000000000000..551b36a087cb78c86091914860f23faa46323eae
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u.h
@@ -0,0 +1,73 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_chebyshev_polynomial_u_ops.h>
+
+namespace at {
+
+
+// aten::special_chebyshev_polynomial_u(Tensor x, Tensor n) -> Tensor
+inline at::Tensor special_chebyshev_polynomial_u(const at::Tensor & x, const at::Tensor & n) {
+    return at::_ops::special_chebyshev_polynomial_u::call(x, n);
+}
+
+// aten::special_chebyshev_polynomial_u.x_scalar(Scalar x, Tensor n) -> Tensor
+inline at::Tensor special_chebyshev_polynomial_u(const at::Scalar & x, const at::Tensor & n) {
+    return at::_ops::special_chebyshev_polynomial_u_x_scalar::call(x, n);
+}
+
+// aten::special_chebyshev_polynomial_u.n_scalar(Tensor x, Scalar n) -> Tensor
+inline at::Tensor special_chebyshev_polynomial_u(const at::Tensor & x, const at::Scalar & n) {
+    return at::_ops::special_chebyshev_polynomial_u_n_scalar::call(x, n);
+}
+
+// aten::special_chebyshev_polynomial_u.out(Tensor x, Tensor n, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_chebyshev_polynomial_u_out(at::Tensor & out, const at::Tensor & x, const at::Tensor & n) {
+    return at::_ops::special_chebyshev_polynomial_u_out::call(x, n, out);
+}
+// aten::special_chebyshev_polynomial_u.out(Tensor x, Tensor n, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_chebyshev_polynomial_u_outf(const at::Tensor & x, const at::Tensor & n, at::Tensor & out) {
+    return at::_ops::special_chebyshev_polynomial_u_out::call(x, n, out);
+}
+
+// aten::special_chebyshev_polynomial_u.x_scalar_out(Scalar x, Tensor n, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_chebyshev_polynomial_u_out(at::Tensor & out, const at::Scalar & x, const at::Tensor & n) {
+    return at::_ops::special_chebyshev_polynomial_u_x_scalar_out::call(x, n, out);
+}
+// aten::special_chebyshev_polynomial_u.x_scalar_out(Scalar x, Tensor n, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_chebyshev_polynomial_u_outf(const at::Scalar & x, const at::Tensor & n, at::Tensor & out) {
+    return at::_ops::special_chebyshev_polynomial_u_x_scalar_out::call(x, n, out);
+}
+
+// aten::special_chebyshev_polynomial_u.n_scalar_out(Tensor x, Scalar n, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_chebyshev_polynomial_u_out(at::Tensor & out, const at::Tensor & x, const at::Scalar & n) {
+    return at::_ops::special_chebyshev_polynomial_u_n_scalar_out::call(x, n, out);
+}
+// aten::special_chebyshev_polynomial_u.n_scalar_out(Tensor x, Scalar n, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_chebyshev_polynomial_u_outf(const at::Tensor & x, const at::Scalar & n, at::Tensor & out) {
+    return at::_ops::special_chebyshev_polynomial_u_n_scalar_out::call(x, n, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..069c0803ed1b4b8904ca2dcbaf56c1f1fe721122
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u_native.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_chebyshev_polynomial_u_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_special_chebyshev_polynomial_u_out : public at::meta::structured_special_chebyshev_polynomial_u {
+void impl(const at::Tensor & x, const at::Tensor & n, const at::Tensor & out);
+};
+TORCH_API at::Tensor special_chebyshev_polynomial_u(const at::Scalar & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_chebyshev_polynomial_u_out(const at::Scalar & x, const at::Tensor & n, at::Tensor & out);
+TORCH_API at::Tensor special_chebyshev_polynomial_u(const at::Tensor & x, const at::Scalar & n);
+TORCH_API at::Tensor & special_chebyshev_polynomial_u_out(const at::Tensor & x, const at::Scalar & n, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..fc845849b82f2ce4bea254e8a1111f4bc4902138
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_u_ops.h
@@ -0,0 +1,89 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 const char* name = "aten::special_chebyshev_polynomial_u";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* 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 const char* name = "aten::special_chebyshev_polynomial_u";
+  static constexpr const char* overload_name = "x_scalar";
+  static constexpr const char* 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 const char* name = "aten::special_chebyshev_polynomial_u";
+  static constexpr const char* overload_name = "n_scalar";
+  static constexpr const char* 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 const char* name = "aten::special_chebyshev_polynomial_u";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* 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 const char* name = "aten::special_chebyshev_polynomial_u";
+  static constexpr const char* overload_name = "x_scalar_out";
+  static constexpr const char* 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 const char* name = "aten::special_chebyshev_polynomial_u";
+  static constexpr const char* overload_name = "n_scalar_out";
+  static constexpr const char* 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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_w_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_w_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..422508f18257bb6420c3e4d373e1805c7c85fb95
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_w_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_chebyshev_polynomial_w(const at::Scalar & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_chebyshev_polynomial_w_out(at::Tensor & out, const at::Scalar & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_chebyshev_polynomial_w_outf(const at::Scalar & x, const at::Tensor & n, at::Tensor & out);
+TORCH_API at::Tensor special_chebyshev_polynomial_w(const at::Tensor & x, const at::Scalar & n);
+TORCH_API at::Tensor & special_chebyshev_polynomial_w_out(at::Tensor & out, const at::Tensor & x, const at::Scalar & n);
+TORCH_API at::Tensor & special_chebyshev_polynomial_w_outf(const at::Tensor & x, const at::Scalar & n, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_w_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_w_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..44259d6d39a43138b66bf35603bcc98e4e1955bd
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_chebyshev_polynomial_w_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_chebyshev_polynomial_w(const at::Tensor & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_chebyshev_polynomial_w_out(at::Tensor & out, const at::Tensor & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_chebyshev_polynomial_w_outf(const at::Tensor & x, const at::Tensor & n, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erf.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erf.h
new file mode 100644
index 0000000000000000000000000000000000000000..e914143f78369923612869201599f6b2c364e0a1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erf.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_erf_ops.h>
+
+namespace at {
+
+
+// aten::special_erf(Tensor self) -> Tensor
+inline at::Tensor special_erf(const at::Tensor & self) {
+    return at::_ops::special_erf::call(self);
+}
+
+// aten::special_erf.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_erf_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::special_erf_out::call(self, out);
+}
+// aten::special_erf.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_erf_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::special_erf_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erf_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erf_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8e305a3cce5ed12f41fea6a78aa0a94ab536dae7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erf_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_erf(const at::Tensor & self);
+TORCH_API at::Tensor & special_erf_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & special_erf_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erf_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erf_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..894999791a8aa14d105d108e97af3047bc06a842
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erf_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_erf {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_erf";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "special_erf(Tensor self) -> Tensor";
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+struct TORCH_API special_erf_out {
+  using schema = at::Tensor & (const at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_erf";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "special_erf.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erfc_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erfc_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9e41e54e4990dc2fec69d9c62d880d0a83b96e11
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_erfc_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_erfc(const at::Tensor & self);
+TORCH_API at::Tensor & special_erfc_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & special_erfc_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_exp2.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_exp2.h
new file mode 100644
index 0000000000000000000000000000000000000000..84aa565961ca2e44f739d72d49c5cfca7215914f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_exp2.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_exp2_ops.h>
+
+namespace at {
+
+
+// aten::special_exp2(Tensor self) -> Tensor
+inline at::Tensor special_exp2(const at::Tensor & self) {
+    return at::_ops::special_exp2::call(self);
+}
+
+// aten::special_exp2.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_exp2_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::special_exp2_out::call(self, out);
+}
+// aten::special_exp2.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_exp2_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::special_exp2_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_hermite_polynomial_h_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_hermite_polynomial_h_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..70b22e7bddffba3a9a4db3e49c697bb85254447e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_hermite_polynomial_h_ops.h
@@ -0,0 +1,89 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_hermite_polynomial_h {
+  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 const char* name = "aten::special_hermite_polynomial_h";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "special_hermite_polynomial_h(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_hermite_polynomial_h_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 const char* name = "aten::special_hermite_polynomial_h";
+  static constexpr const char* overload_name = "x_scalar";
+  static constexpr const char* schema_str = "special_hermite_polynomial_h.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_hermite_polynomial_h_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 const char* name = "aten::special_hermite_polynomial_h";
+  static constexpr const char* overload_name = "n_scalar";
+  static constexpr const char* schema_str = "special_hermite_polynomial_h.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_hermite_polynomial_h_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 const char* name = "aten::special_hermite_polynomial_h";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "special_hermite_polynomial_h.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_hermite_polynomial_h_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 const char* name = "aten::special_hermite_polynomial_h";
+  static constexpr const char* overload_name = "x_scalar_out";
+  static constexpr const char* schema_str = "special_hermite_polynomial_h.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_hermite_polynomial_h_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 const char* name = "aten::special_hermite_polynomial_h";
+  static constexpr const char* overload_name = "n_scalar_out";
+  static constexpr const char* schema_str = "special_hermite_polynomial_h.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_hermite_polynomial_he_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_hermite_polynomial_he_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..a5038c0650a53bd35d1042d9a9acb89725c44d4d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_hermite_polynomial_he_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_hermite_polynomial_he(const at::Tensor & x, const at::Tensor & n);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_hermite_polynomial_he_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_hermite_polynomial_he_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..51d5d9b1e80ee14255c24976b79a6ae5f5714055
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_hermite_polynomial_he_ops.h
@@ -0,0 +1,89 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_hermite_polynomial_he {
+  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 const char* name = "aten::special_hermite_polynomial_he";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "special_hermite_polynomial_he(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_hermite_polynomial_he_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 const char* name = "aten::special_hermite_polynomial_he";
+  static constexpr const char* overload_name = "x_scalar";
+  static constexpr const char* schema_str = "special_hermite_polynomial_he.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_hermite_polynomial_he_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 const char* name = "aten::special_hermite_polynomial_he";
+  static constexpr const char* overload_name = "n_scalar";
+  static constexpr const char* schema_str = "special_hermite_polynomial_he.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_hermite_polynomial_he_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 const char* name = "aten::special_hermite_polynomial_he";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "special_hermite_polynomial_he.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_hermite_polynomial_he_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 const char* name = "aten::special_hermite_polynomial_he";
+  static constexpr const char* overload_name = "x_scalar_out";
+  static constexpr const char* schema_str = "special_hermite_polynomial_he.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_hermite_polynomial_he_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 const char* name = "aten::special_hermite_polynomial_he";
+  static constexpr const char* overload_name = "n_scalar_out";
+  static constexpr const char* schema_str = "special_hermite_polynomial_he.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i0_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i0_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..f9fac3d3a08e113776ba11bd43bec05ff4a48392
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i0_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_i0(const at::Tensor & self);
+TORCH_API at::Tensor & special_i0_out(const at::Tensor & self, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i0e_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i0e_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..a4f134874d595c9764272d0cc67a11e36015c998
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i0e_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_i0e {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_i0e";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "special_i0e(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 special_i0e_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 const char* name = "aten::special_i0e";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "special_i0e.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i1_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i1_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..336dc83ee0bad68633005caa35fc87fb8f3d0d07
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i1_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_i1(const at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i1_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i1_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..65bcbacf8354e5c0f0259ac5e50220ed8d5a53af
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i1_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_i1(const at::Tensor & self);
+TORCH_API at::Tensor & special_i1_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & special_i1_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i1_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i1_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..9a89fefe7934ea65827f8477a505b59edfe1b90e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_i1_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_i1 {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_i1";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "special_i1(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 special_i1_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 const char* name = "aten::special_i1";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "special_i1.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..67f76ec651ca94425163c2ed0e0a70cac6fc82a0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_laguerre_polynomial_l(const at::Scalar & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_laguerre_polynomial_l_out(at::Tensor & out, const at::Scalar & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_laguerre_polynomial_l_outf(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(at::Tensor & out, const at::Tensor & x, const at::Scalar & n);
+TORCH_API at::Tensor & special_laguerre_polynomial_l_outf(const at::Tensor & x, const at::Scalar & n, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9827f112cca9ef56bf03408444e6793fec9fc1a3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_laguerre_polynomial_l(const at::Tensor & x, const at::Tensor & n);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3ba7996bd55b35790d38c373c07e55a820e740f5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_laguerre_polynomial_l_cpu_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_laguerre_polynomial_l(const at::Tensor & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_laguerre_polynomial_l_out(at::Tensor & out, const at::Tensor & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_laguerre_polynomial_l_outf(const at::Tensor & x, const at::Tensor & n, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_legendre_polynomial_p_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_legendre_polynomial_p_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..7e46187f5d93dc54699a29e0fe94d99eea7397cc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_legendre_polynomial_p_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_legendre_polynomial_p(const at::Tensor & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_legendre_polynomial_p_out(at::Tensor & out, const at::Tensor & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_legendre_polynomial_p_outf(const at::Tensor & x, const at::Tensor & n, at::Tensor & out);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_ndtr_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_ndtr_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..0e20bab4ca47441d1c7b1c0ad0fe5d8afe42d625
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_ndtr_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_log_ndtr_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_special_log_ndtr_out : public at::meta::structured_special_log_ndtr {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_ndtr_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_ndtr_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..f97949f06ed009f09a4404000d5141096bc1aad5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_ndtr_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_log_ndtr {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_log_ndtr";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "special_log_ndtr(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 special_log_ndtr_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 const char* name = "aten::special_log_ndtr";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "special_log_ndtr.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_softmax.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_softmax.h
new file mode 100644
index 0000000000000000000000000000000000000000..833af9ebb68da67f916c94479710425ba1e86309
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_softmax.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_log_softmax_ops.h>
+
+namespace at {
+
+
+// aten::special_log_softmax(Tensor self, int dim, *, ScalarType? dtype=None) -> Tensor
+inline at::Tensor special_log_softmax(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt) {
+    return at::_ops::special_log_softmax::call(self, dim, dtype);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_softmax_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_softmax_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..97af3eeafa7db6d84325184aff662570877aafcb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_softmax_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_log_softmax(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_softmax_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_softmax_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..0943a56bea31f0dfc2f3425e8525fce957070ae3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_log_softmax_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_log_softmax(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_logit.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_logit.h
new file mode 100644
index 0000000000000000000000000000000000000000..cf5ec9a42fb2273ad889ba598e07c79a9a6f3fa6
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_logit.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_logit_ops.h>
+
+namespace at {
+
+
+// aten::special_logit(Tensor self, float? eps=None) -> Tensor
+inline at::Tensor special_logit(const at::Tensor & self, ::std::optional<double> eps=::std::nullopt) {
+    return at::_ops::special_logit::call(self, eps);
+}
+
+// aten::special_logit.out(Tensor self, float? eps=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_logit_out(at::Tensor & out, const at::Tensor & self, ::std::optional<double> eps=::std::nullopt) {
+    return at::_ops::special_logit_out::call(self, eps, out);
+}
+// aten::special_logit.out(Tensor self, float? eps=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_logit_outf(const at::Tensor & self, ::std::optional<double> eps, at::Tensor & out) {
+    return at::_ops::special_logit_out::call(self, eps, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_logit_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_logit_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..94e478dc76fc5467842069d870475ffa86c12ffe
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_logit_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_logit {
+  using schema = at::Tensor (const at::Tensor &, ::std::optional<double>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_logit";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "special_logit(Tensor self, float? eps=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, ::std::optional<double> eps);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<double> eps);
+};
+
+struct TORCH_API special_logit_out {
+  using schema = at::Tensor & (const at::Tensor &, ::std::optional<double>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_logit";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "special_logit.out(Tensor self, float? eps=None, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, ::std::optional<double> eps, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, ::std::optional<double> eps, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_i0.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_i0.h
new file mode 100644
index 0000000000000000000000000000000000000000..e0f2c6d9372b10b5b405f593989ef140c40c6219
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_i0.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_modified_bessel_i0_ops.h>
+
+namespace at {
+
+
+// aten::special_modified_bessel_i0(Tensor self) -> Tensor
+inline at::Tensor special_modified_bessel_i0(const at::Tensor & self) {
+    return at::_ops::special_modified_bessel_i0::call(self);
+}
+
+// aten::special_modified_bessel_i0.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_modified_bessel_i0_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::special_modified_bessel_i0_out::call(self, out);
+}
+// aten::special_modified_bessel_i0.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_modified_bessel_i0_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::special_modified_bessel_i0_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_i0_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_i0_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..be178bfc320d032319354943edb18d0ec39532e7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_i0_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_modified_bessel_i0_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_special_modified_bessel_i0_out : public at::meta::structured_special_modified_bessel_i0 {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_i1_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_i1_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..94fecc219ffc5127d42aaa919647dd552fcb1e7f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_i1_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_modified_bessel_i1(const at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_k1_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_k1_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..3589a9525843b05f3f1ba7824314d22e376c8a74
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_modified_bessel_k1_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_special_modified_bessel_k1 : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_multigammaln.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_multigammaln.h
new file mode 100644
index 0000000000000000000000000000000000000000..92260c7f165326da3b2a501f43cac788471dd291
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_multigammaln.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_multigammaln_ops.h>
+
+namespace at {
+
+
+// aten::special_multigammaln(Tensor self, int p) -> Tensor
+inline at::Tensor special_multigammaln(const at::Tensor & self, int64_t p) {
+    return at::_ops::special_multigammaln::call(self, p);
+}
+
+// aten::special_multigammaln.out(Tensor self, int p, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_multigammaln_out(at::Tensor & out, const at::Tensor & self, int64_t p) {
+    return at::_ops::special_multigammaln_out::call(self, p, out);
+}
+// aten::special_multigammaln.out(Tensor self, int p, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_multigammaln_outf(const at::Tensor & self, int64_t p, at::Tensor & out) {
+    return at::_ops::special_multigammaln_out::call(self, p, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_multigammaln_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_multigammaln_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f5d8465b1e53f2ba2cc91358a52a774c76c57c69
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_multigammaln_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_multigammaln(const at::Tensor & self, int64_t p);
+TORCH_API at::Tensor & special_multigammaln_out(at::Tensor & out, const at::Tensor & self, int64_t p);
+TORCH_API at::Tensor & special_multigammaln_outf(const at::Tensor & self, int64_t p, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_ndtri_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_ndtri_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..e6e71c76825e33e1c67c023f38e5a3494345ad1b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_ndtri_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_special_ndtri : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_ndtri_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_ndtri_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..34afb6bd8b1305b1a6a2b53b23c931464a8b8d52
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_ndtri_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_ndtri {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_ndtri";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "special_ndtri(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 special_ndtri_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 const char* name = "aten::special_ndtri";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "special_ndtri.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_polygamma.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_polygamma.h
new file mode 100644
index 0000000000000000000000000000000000000000..a3fe18894736d9c27fdbff10e855da6fed286ce7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_polygamma.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_polygamma_ops.h>
+
+namespace at {
+
+
+// aten::special_polygamma(int n, Tensor self) -> Tensor
+inline at::Tensor special_polygamma(int64_t n, const at::Tensor & self) {
+    return at::_ops::special_polygamma::call(n, self);
+}
+
+// aten::special_polygamma.out(int n, Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_polygamma_out(at::Tensor & out, int64_t n, const at::Tensor & self) {
+    return at::_ops::special_polygamma_out::call(n, self, out);
+}
+// aten::special_polygamma.out(int n, Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_polygamma_outf(int64_t n, const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::special_polygamma_out::call(n, self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_psi.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_psi.h
new file mode 100644
index 0000000000000000000000000000000000000000..6816ac1db3f471f0e726848df575f70a3587110c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_psi.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_psi_ops.h>
+
+namespace at {
+
+
+// aten::special_psi(Tensor self) -> Tensor
+inline at::Tensor special_psi(const at::Tensor & self) {
+    return at::_ops::special_psi::call(self);
+}
+
+// aten::special_psi.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_psi_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::special_psi_out::call(self, out);
+}
+// aten::special_psi.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_psi_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::special_psi_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_round.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_round.h
new file mode 100644
index 0000000000000000000000000000000000000000..a051b45704f6b159a9df14ba2c6720d315fb2ce7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_round.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_round_ops.h>
+
+namespace at {
+
+
+// aten::special_round(Tensor self, *, int decimals=0) -> Tensor
+inline at::Tensor special_round(const at::Tensor & self, int64_t decimals=0) {
+    return at::_ops::special_round::call(self, decimals);
+}
+
+// aten::special_round.out(Tensor self, *, int decimals=0, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_round_out(at::Tensor & out, const at::Tensor & self, int64_t decimals=0) {
+    return at::_ops::special_round_out::call(self, decimals, out);
+}
+// aten::special_round.out(Tensor self, *, int decimals=0, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_round_outf(const at::Tensor & self, int64_t decimals, at::Tensor & out) {
+    return at::_ops::special_round_out::call(self, decimals, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_round_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_round_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..fc512ab7b9e1115802178f5599407c4c995f8148
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_round_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_round {
+  using schema = at::Tensor (const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_round";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "special_round(Tensor self, *, int decimals=0) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t decimals);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t decimals);
+};
+
+struct TORCH_API special_round_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_round";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "special_round.out(Tensor self, *, int decimals=0, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, int64_t decimals, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t decimals, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_scaled_modified_bessel_k0.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_scaled_modified_bessel_k0.h
new file mode 100644
index 0000000000000000000000000000000000000000..d5a29910019b3206ac2b8cc8462e308345b95ff5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_scaled_modified_bessel_k0.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_scaled_modified_bessel_k0_ops.h>
+
+namespace at {
+
+
+// aten::special_scaled_modified_bessel_k0(Tensor x) -> Tensor
+inline at::Tensor special_scaled_modified_bessel_k0(const at::Tensor & x) {
+    return at::_ops::special_scaled_modified_bessel_k0::call(x);
+}
+
+// aten::special_scaled_modified_bessel_k0.out(Tensor x, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_scaled_modified_bessel_k0_out(at::Tensor & out, const at::Tensor & x) {
+    return at::_ops::special_scaled_modified_bessel_k0_out::call(x, out);
+}
+// aten::special_scaled_modified_bessel_k0.out(Tensor x, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_scaled_modified_bessel_k0_outf(const at::Tensor & x, at::Tensor & out) {
+    return at::_ops::special_scaled_modified_bessel_k0_out::call(x, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_scaled_modified_bessel_k0_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_scaled_modified_bessel_k0_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3ff7d493502335d855a5f3f6ea77bc473d1cccfb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_scaled_modified_bessel_k0_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_scaled_modified_bessel_k0(const at::Tensor & x);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_scaled_modified_bessel_k1.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_scaled_modified_bessel_k1.h
new file mode 100644
index 0000000000000000000000000000000000000000..f715313216c208d83f3775d1fe991dd5ec30425d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_scaled_modified_bessel_k1.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/special_scaled_modified_bessel_k1_ops.h>
+
+namespace at {
+
+
+// aten::special_scaled_modified_bessel_k1(Tensor x) -> Tensor
+inline at::Tensor special_scaled_modified_bessel_k1(const at::Tensor & x) {
+    return at::_ops::special_scaled_modified_bessel_k1::call(x);
+}
+
+// aten::special_scaled_modified_bessel_k1.out(Tensor x, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_scaled_modified_bessel_k1_out(at::Tensor & out, const at::Tensor & x) {
+    return at::_ops::special_scaled_modified_bessel_k1_out::call(x, out);
+}
+// aten::special_scaled_modified_bessel_k1.out(Tensor x, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & special_scaled_modified_bessel_k1_outf(const at::Tensor & x, at::Tensor & out) {
+    return at::_ops::special_scaled_modified_bessel_k1_out::call(x, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_t_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_t_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..dd69f9818b57709893a72f4040fa2f22ccc5cd5a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_t_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_shifted_chebyshev_polynomial_t(const at::Tensor & x, const at::Tensor & n);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_t_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_t_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..115a135cf06e1115de155a087ea5002aaef8143d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_t_native.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_shifted_chebyshev_polynomial_t_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_special_shifted_chebyshev_polynomial_t_out : public at::meta::structured_special_shifted_chebyshev_polynomial_t {
+void impl(const at::Tensor & x, const at::Tensor & n, const at::Tensor & out);
+};
+TORCH_API at::Tensor special_shifted_chebyshev_polynomial_t(const at::Scalar & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_shifted_chebyshev_polynomial_t_out(const at::Scalar & x, const at::Tensor & n, at::Tensor & out);
+TORCH_API at::Tensor special_shifted_chebyshev_polynomial_t(const at::Tensor & x, const at::Scalar & n);
+TORCH_API at::Tensor & special_shifted_chebyshev_polynomial_t_out(const at::Tensor & x, const at::Scalar & n, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_u_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_u_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..cc2b5217b30d90ed11e15b36cd04a5db4d437f80
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_u_native.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_shifted_chebyshev_polynomial_u_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_special_shifted_chebyshev_polynomial_u_out : public at::meta::structured_special_shifted_chebyshev_polynomial_u {
+void impl(const at::Tensor & x, const at::Tensor & n, const at::Tensor & out);
+};
+TORCH_API at::Tensor special_shifted_chebyshev_polynomial_u(const at::Scalar & x, const at::Tensor & n);
+TORCH_API at::Tensor & special_shifted_chebyshev_polynomial_u_out(const at::Scalar & x, const at::Tensor & n, at::Tensor & out);
+TORCH_API at::Tensor special_shifted_chebyshev_polynomial_u(const at::Tensor & x, const at::Scalar & n);
+TORCH_API at::Tensor & special_shifted_chebyshev_polynomial_u_out(const at::Tensor & x, const at::Scalar & n, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_v_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_v_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..faed2a57e16734670e39f545835e59805a11ea40
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_shifted_chebyshev_polynomial_v_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_special_shifted_chebyshev_polynomial_v : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & x, const at::Tensor & n);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_softmax_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_softmax_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b9b5f21c576adb742298623a21573d0dba8abc50
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_softmax_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_softmax(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_softmax_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_softmax_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..e618b411e8e205d4d46a665cfeb80c46b2ebd5b4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_softmax_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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_softmax {
+  using schema = at::Tensor (const at::Tensor &, int64_t, ::std::optional<at::ScalarType>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::special_softmax";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "special_softmax(Tensor self, int dim, ScalarType? dtype=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_xlog1py_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_xlog1py_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e2b4d02ac8a17701d0bd05c0b86d72ebb7a6d4c8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_xlog1py_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_xlog1py(const at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_xlog1py_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_xlog1py_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..ee990de61e862c9297ccfb8979c9790c6c28deb3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_xlog1py_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_special_xlog1py : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & self, const at::Tensor & other);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_zeta_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_zeta_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..df455651be1915e78672fc427ab99bdc20537822
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_zeta_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_zeta(const at::Scalar & self, const at::Tensor & other);
+TORCH_API at::Tensor & special_zeta_out(at::Tensor & out, const at::Scalar & self, const at::Tensor & other);
+TORCH_API at::Tensor & special_zeta_outf(const at::Scalar & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor special_zeta(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & special_zeta_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & special_zeta_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_zeta_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_zeta_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..28c37b6b48b6dfd7fe0b4fb1f07ff2c140c75b54
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/special_zeta_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 special_zeta(const at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/split_copy_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/split_copy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..7dee4243ec57cb3c7c6604f774fa29b99e37e285
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/split_copy_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void split_copy_Tensor_out(const at::Tensor & self, int64_t split_size, int64_t dim, at::TensorList out);
+TORCH_API ::std::vector<at::Tensor> split_copy_Tensor_symint(const at::Tensor & self, c10::SymInt split_size, int64_t dim=0);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/split_with_sizes_copy_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/split_with_sizes_copy_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6c401501470458a7a94832fdc8a38d9b8729c9c2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/split_with_sizes_copy_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 void split_with_sizes_copy_out(at::TensorList out, const at::Tensor & self, at::IntArrayRef split_sizes, int64_t dim=0);
+TORCH_API void split_with_sizes_copy_outf(const at::Tensor & self, at::IntArrayRef split_sizes, int64_t dim, at::TensorList out);
+TORCH_API void split_with_sizes_copy_symint_out(at::TensorList out, const at::Tensor & self, c10::SymIntArrayRef split_sizes, int64_t dim=0);
+TORCH_API void split_with_sizes_copy_symint_outf(const at::Tensor & self, c10::SymIntArrayRef split_sizes, int64_t dim, at::TensorList out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sqrt_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sqrt_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..47f720bd4ba2dc0a15f3d0e761a72cc2f974e520
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sqrt_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sqrt(const at::Tensor & self);
+TORCH_API at::Tensor & sqrt_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sqrt_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sqrt_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..fbcd43792cdcf233d16dbf4836c42b69499a94fb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sqrt_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_sqrt : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/square_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/square_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..e0894cd9c197045848ae054d8188b0a3b69b8b5c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/square_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 square(const at::Tensor & self);
+TORCH_API at::Tensor & square_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & square_(at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/squeeze_copy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/squeeze_copy.h
new file mode 100644
index 0000000000000000000000000000000000000000..5c0f72f14e1d20159c04c99b02677becd18bc78a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/squeeze_copy.h
@@ -0,0 +1,73 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/squeeze_copy_ops.h>
+
+namespace at {
+
+
+// aten::squeeze_copy(Tensor self) -> Tensor
+inline at::Tensor squeeze_copy(const at::Tensor & self) {
+    return at::_ops::squeeze_copy::call(self);
+}
+
+// aten::squeeze_copy.dim(Tensor self, int dim) -> Tensor
+inline at::Tensor squeeze_copy(const at::Tensor & self, int64_t dim) {
+    return at::_ops::squeeze_copy_dim::call(self, dim);
+}
+
+// aten::squeeze_copy.dims(Tensor self, int[] dim) -> Tensor
+inline at::Tensor squeeze_copy(const at::Tensor & self, at::IntArrayRef dim) {
+    return at::_ops::squeeze_copy_dims::call(self, dim);
+}
+
+// aten::squeeze_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & squeeze_copy_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::squeeze_copy_out::call(self, out);
+}
+// aten::squeeze_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & squeeze_copy_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::squeeze_copy_out::call(self, out);
+}
+
+// aten::squeeze_copy.dim_out(Tensor self, int dim, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & squeeze_copy_out(at::Tensor & out, const at::Tensor & self, int64_t dim) {
+    return at::_ops::squeeze_copy_dim_out::call(self, dim, out);
+}
+// aten::squeeze_copy.dim_out(Tensor self, int dim, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & squeeze_copy_outf(const at::Tensor & self, int64_t dim, at::Tensor & out) {
+    return at::_ops::squeeze_copy_dim_out::call(self, dim, out);
+}
+
+// aten::squeeze_copy.dims_out(Tensor self, int[] dim, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & squeeze_copy_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef dim) {
+    return at::_ops::squeeze_copy_dims_out::call(self, dim, out);
+}
+// aten::squeeze_copy.dims_out(Tensor self, int[] dim, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & squeeze_copy_outf(const at::Tensor & self, at::IntArrayRef dim, at::Tensor & out) {
+    return at::_ops::squeeze_copy_dims_out::call(self, dim, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/squeeze_copy_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/squeeze_copy_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..549c282cf814994c9e8527cf7725b7a418fd0d00
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/squeeze_copy_native.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & squeeze_copy_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor squeeze_copy(const at::Tensor & self);
+TORCH_API at::Tensor & squeeze_copy_dim_out(const at::Tensor & self, int64_t dim, at::Tensor & out);
+TORCH_API at::Tensor squeeze_copy_dim(const at::Tensor & self, int64_t dim);
+TORCH_API at::Tensor & squeeze_copy_dims_out(const at::Tensor & self, at::IntArrayRef dim, at::Tensor & out);
+TORCH_API at::Tensor squeeze_copy_dims(const at::Tensor & self, at::IntArrayRef dim);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/squeeze_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/squeeze_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..f1ac1fa61468f0b667435b2e48390ac1bb87ee0e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/squeeze_ops.h
@@ -0,0 +1,111 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 squeeze {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::squeeze";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "squeeze(Tensor(a) self) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+struct TORCH_API squeeze_dim {
+  using schema = at::Tensor (const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::squeeze";
+  static constexpr const char* overload_name = "dim";
+  static constexpr const char* schema_str = "squeeze.dim(Tensor(a) self, int dim) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, int64_t dim);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim);
+};
+
+struct TORCH_API squeeze_dimname {
+  using schema = at::Tensor (const at::Tensor &, at::Dimname);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::squeeze";
+  static constexpr const char* overload_name = "dimname";
+  static constexpr const char* schema_str = "squeeze.dimname(Tensor(a) self, Dimname dim) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, at::Dimname dim);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Dimname dim);
+};
+
+struct TORCH_API squeeze_dims {
+  using schema = at::Tensor (const at::Tensor &, at::IntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::squeeze";
+  static constexpr const char* overload_name = "dims";
+  static constexpr const char* schema_str = "squeeze.dims(Tensor(a) self, int[] dim) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, at::IntArrayRef dim);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef dim);
+};
+
+struct TORCH_API squeeze_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::squeeze_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "squeeze_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API squeeze__dim {
+  using schema = at::Tensor & (at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::squeeze_";
+  static constexpr const char* overload_name = "dim";
+  static constexpr const char* schema_str = "squeeze_.dim(Tensor(a!) self, int dim) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, int64_t dim);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, int64_t dim);
+};
+
+struct TORCH_API squeeze__dims {
+  using schema = at::Tensor & (at::Tensor &, at::IntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::squeeze_";
+  static constexpr const char* overload_name = "dims";
+  static constexpr const char* schema_str = "squeeze_.dims(Tensor(a!) self, int[] dim) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, at::IntArrayRef dim);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, at::IntArrayRef dim);
+};
+
+struct TORCH_API squeeze__dimname {
+  using schema = at::Tensor & (at::Tensor &, at::Dimname);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::squeeze_";
+  static constexpr const char* overload_name = "dimname";
+  static constexpr const char* schema_str = "squeeze_.dimname(Tensor(a!) self, Dimname dim) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, at::Dimname dim);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, at::Dimname dim);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/std_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/std_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..be492b8cd0c77a9be554fe721ad9c72555ad58bc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/std_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,37 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 std(const at::Tensor & self, bool unbiased);
+TORCH_API at::Tensor std(const at::Tensor & self, at::OptionalIntArrayRef dim, bool unbiased, bool keepdim=false);
+TORCH_API at::Tensor & std_out(at::Tensor & out, const at::Tensor & self, at::OptionalIntArrayRef dim, bool unbiased, bool keepdim=false);
+TORCH_API at::Tensor & std_outf(const at::Tensor & self, at::OptionalIntArrayRef dim, bool unbiased, bool keepdim, at::Tensor & out);
+TORCH_API at::Tensor std(const at::Tensor & self, at::DimnameList dim, bool unbiased, bool keepdim=false);
+TORCH_API at::Tensor & std_out(at::Tensor & out, const at::Tensor & self, at::DimnameList dim, bool unbiased, bool keepdim=false);
+TORCH_API at::Tensor & std_outf(const at::Tensor & self, at::DimnameList dim, bool unbiased, bool keepdim, at::Tensor & out);
+TORCH_API at::Tensor std(const at::Tensor & self, at::DimnameList dim, const ::std::optional<at::Scalar> & correction=::std::nullopt, bool keepdim=false);
+TORCH_API at::Tensor & std_out(at::Tensor & out, const at::Tensor & self, at::DimnameList dim, const ::std::optional<at::Scalar> & correction=::std::nullopt, bool keepdim=false);
+TORCH_API at::Tensor & std_outf(const at::Tensor & self, at::DimnameList dim, const ::std::optional<at::Scalar> & correction, bool keepdim, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/std_mean_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/std_mean_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3335993bf1e13254455d804c61eb722e2beed5d3
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/std_mean_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 ::std::tuple<at::Tensor,at::Tensor> std_mean(const at::Tensor & self, bool unbiased);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> std_mean(const at::Tensor & self, at::OptionalIntArrayRef dim, bool unbiased, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> std_mean(const at::Tensor & self, at::DimnameList dim, bool unbiased, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> std_mean(const at::Tensor & self, at::DimnameList dim, const ::std::optional<at::Scalar> & correction=::std::nullopt, bool keepdim=false);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sub_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sub_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..bc2183f01020bd3814ded0f86040a5e2c168423c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sub_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 sub(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & sub_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & sub_outf(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & sub_(at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/subtract_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/subtract_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..ef3728b2d358e8865a1787372f7d23d26d91b27d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/subtract_ops.h
@@ -0,0 +1,78 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 subtract_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::subtract";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "subtract.out(Tensor self, Tensor other, *, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out);
+};
+
+struct TORCH_API subtract_Tensor {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::subtract";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "subtract.Tensor(Tensor self, Tensor other, *, Scalar alpha=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha);
+};
+
+struct TORCH_API subtract__Tensor {
+  using schema = at::Tensor & (at::Tensor &, const at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::subtract_";
+  static constexpr const char* overload_name = "Tensor";
+  static constexpr const char* schema_str = "subtract_.Tensor(Tensor(a!) self, Tensor other, *, Scalar alpha=1) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha);
+};
+
+struct TORCH_API subtract_Scalar {
+  using schema = at::Tensor (const at::Tensor &, const at::Scalar &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::subtract";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "subtract.Scalar(Tensor self, Scalar other, Scalar alpha=1) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha);
+};
+
+struct TORCH_API subtract__Scalar {
+  using schema = at::Tensor & (at::Tensor &, const at::Scalar &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::subtract_";
+  static constexpr const char* overload_name = "Scalar";
+  static constexpr const char* schema_str = "subtract_.Scalar(Tensor(a!) self, Scalar other, Scalar alpha=1) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sum_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sum_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..5da60f83c2bd199a773ed7dc80e6b143fd6b2a91
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sum_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_sum_dim_IntList : public at::impl::MetaBase {
+
+
+    void meta(const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim, ::std::optional<at::ScalarType> dtype);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/svd_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/svd_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..6e8a8a1c516d973f7dfd2b10b9fdc747bade7f39
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/svd_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> svd(const at::Tensor & self, bool some=true, bool compute_uv=true);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> svd_out(const at::Tensor & self, bool some, bool compute_uv, at::Tensor & U, at::Tensor & S, at::Tensor & V);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_constrain_range.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_constrain_range.h
new file mode 100644
index 0000000000000000000000000000000000000000..363df8bd1957b53a69b12d2cec39a345955c8674
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_constrain_range.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/sym_constrain_range_ops.h>
+
+namespace at {
+
+
+// aten::sym_constrain_range(Scalar size, *, int? min=None, int? max=None) -> ()
+inline void sym_constrain_range(const at::Scalar & size, ::std::optional<int64_t> min=::std::nullopt, ::std::optional<int64_t> max=::std::nullopt) {
+    return at::_ops::sym_constrain_range::call(size, min, max);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_constrain_range_for_size.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_constrain_range_for_size.h
new file mode 100644
index 0000000000000000000000000000000000000000..86012db0371d9875ce10251164a29f096e876836
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_constrain_range_for_size.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/sym_constrain_range_for_size_ops.h>
+
+namespace at {
+
+
+// aten::sym_constrain_range_for_size(Scalar size, *, int? min=None, int? max=None) -> ()
+inline void sym_constrain_range_for_size(const at::Scalar & size, ::std::optional<int64_t> min=::std::nullopt, ::std::optional<int64_t> max=::std::nullopt) {
+    return at::_ops::sym_constrain_range_for_size::call(size, min, max);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_numel_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_numel_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..29c405ebfb1d4ea983134d625e0017fb80b8f1ce
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_numel_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 sym_numel {
+  using schema = c10::SymInt (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::sym_numel";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "sym_numel(Tensor self) -> SymInt";
+  static c10::SymInt call(const at::Tensor & self);
+  static c10::SymInt redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_storage_offset_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_storage_offset_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c39b222412080ec9755792bbb84c9acb3426a21c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/sym_storage_offset_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 c10::SymInt sym_storage_offset(const at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/t.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/t.h
new file mode 100644
index 0000000000000000000000000000000000000000..6765b51d9b3f9ca5139d01745ce70c111f087e69
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/t.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#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);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/t_copy_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/t_copy_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b911315bee3a5fb0753c400deb97bd3f4f35dca0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/t_copy_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 t_copy(const at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/take_along_dim_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/take_along_dim_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8986ad0542969969b8a70572023ead1d943b53cc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/take_along_dim_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 take_along_dim(const at::Tensor & self, const at::Tensor & indices, ::std::optional<int64_t> dim=::std::nullopt);
+TORCH_API at::Tensor & take_along_dim_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & indices, ::std::optional<int64_t> dim=::std::nullopt);
+TORCH_API at::Tensor & take_along_dim_outf(const at::Tensor & self, const at::Tensor & indices, ::std::optional<int64_t> dim, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/take_along_dim_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/take_along_dim_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..de2eef0f5e82e32e73d62a6aa516afeaa0cbe667
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/take_along_dim_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 take_along_dim(const at::Tensor & self, const at::Tensor & indices, ::std::optional<int64_t> dim=::std::nullopt);
+TORCH_API at::Tensor & take_along_dim_out(const at::Tensor & self, const at::Tensor & indices, ::std::optional<int64_t> dim, at::Tensor & out);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tan_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tan_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..18ca1873c08ae2f596b60ab1d2c6ad363def4016
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tan_native.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/tan_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_tan_out : public at::meta::structured_tan {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+TORCH_API at::Tensor tan_sparse(const at::Tensor & self);
+TORCH_API at::Tensor & tan_sparse_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & tan_sparse_(at::Tensor & self);
+TORCH_API at::Tensor tan_sparse_csr(const at::Tensor & self);
+TORCH_API at::Tensor & tan_sparse_csr_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & tan_sparse_csr_(at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tanh_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tanh_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..51088afdcdc854caaa31cf6ebf9fb2a71f7ff6ea
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tanh_cuda_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 tanh(const at::Tensor & self);
+TORCH_API at::Tensor & tanh_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & tanh_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & tanh_(at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tanh_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tanh_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..f0b487e75acba5ae2d20230fc8bd9338d4feb841
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tanh_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_tanh : public TensorIteratorBase {
+
+
+    void meta(const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tanh_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tanh_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..219a9bfdefe9d89564ea0967fc9c162668ea6845
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tanh_native.h
@@ -0,0 +1,39 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/tanh_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_tanh_out : public at::meta::structured_tanh {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+TORCH_API at::Tensor NestedTensor_tanh(const at::Tensor & self);
+TORCH_API at::Tensor & NestedTensor_tanh_(at::Tensor & self);
+TORCH_API at::Tensor tanh_sparse(const at::Tensor & self);
+TORCH_API at::Tensor & tanh_sparse_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & tanh_sparse_(at::Tensor & self);
+TORCH_API at::Tensor tanh_sparse_csr(const at::Tensor & self);
+TORCH_API at::Tensor & tanh_sparse_csr_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & tanh_sparse_csr_(at::Tensor & self);
+TORCH_API at::Tensor mkldnn_tanh(const at::Tensor & self);
+TORCH_API at::Tensor & mkldnn_tanh_(at::Tensor & self);
+TORCH_API at::Tensor tanh_quantized_cpu(const at::Tensor & self);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor.h
new file mode 100644
index 0000000000000000000000000000000000000000..548ae03583f03b85af6d8c5934d05406f0aa2ed1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor.h
@@ -0,0 +1,35 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+#include <ATen/core/Tensor.h>
+#include <c10/core/ScalarType.h>
+
+namespace at {
+
+// These functions are defined in ATen/Utils.cpp.
+#define TENSOR(T, S)                                                          \
+  TORCH_API Tensor tensor(ArrayRef<T> values, const TensorOptions& options);  \
+  inline Tensor tensor(                                                       \
+      std::initializer_list<T> values, const TensorOptions& options) {        \
+    return at::tensor(ArrayRef<T>(values), options);                          \
+  }                                                                           \
+  inline Tensor tensor(T value, const TensorOptions& options) {               \
+    return at::tensor(ArrayRef<T>(value), options);                           \
+  }                                                                           \
+  inline Tensor tensor(ArrayRef<T> values) {                                  \
+    return at::tensor(std::move(values), at::dtype(k##S));                    \
+  }                                                                           \
+  inline Tensor tensor(std::initializer_list<T> values) {                     \
+    return at::tensor(ArrayRef<T>(values));                                   \
+  }                                                                           \
+  inline Tensor tensor(T value) {                                             \
+    return at::tensor(ArrayRef<T>(value));                                    \
+  }
+AT_FORALL_SCALAR_TYPES_AND3(Bool, Half, BFloat16, TENSOR)
+AT_FORALL_COMPLEX_TYPES(TENSOR)
+#undef TENSOR
+
+}  // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor_split.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor_split.h
new file mode 100644
index 0000000000000000000000000000000000000000..0a45854b89cbd96b05bf5b74975a3be1202f39a4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor_split.h
@@ -0,0 +1,80 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/tensor_split_ops.h>
+
+namespace at {
+
+
+// aten::tensor_split.sections(Tensor(a -> *) self, SymInt sections, int dim=0) -> Tensor(a)[]
+inline ::std::vector<at::Tensor> tensor_split(const at::Tensor & self, int64_t sections, int64_t dim=0) {
+    return at::_ops::tensor_split_sections::call(self, sections, dim);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::vector<at::Tensor> tensor_split(const at::Tensor & self, int64_t sections, int64_t dim=0) {
+    return at::_ops::tensor_split_sections::call(self, sections, dim);
+  }
+}
+
+// aten::tensor_split.sections(Tensor(a -> *) self, SymInt sections, int dim=0) -> Tensor(a)[]
+inline ::std::vector<at::Tensor> tensor_split_symint(const at::Tensor & self, c10::SymInt sections, int64_t dim=0) {
+    return at::_ops::tensor_split_sections::call(self, sections, dim);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::vector<at::Tensor> tensor_split(const at::Tensor & self, c10::SymInt sections, int64_t dim=0) {
+    return at::_ops::tensor_split_sections::call(self, sections, dim);
+  }
+}
+
+// aten::tensor_split.indices(Tensor(a -> *) self, SymInt[] indices, int dim=0) -> Tensor(a)[]
+inline ::std::vector<at::Tensor> tensor_split(const at::Tensor & self, at::IntArrayRef indices, int64_t dim=0) {
+    return at::_ops::tensor_split_indices::call(self, c10::fromIntArrayRefSlow(indices), dim);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  ::std::vector<at::Tensor> tensor_split(const at::Tensor & self, at::IntArrayRef indices, int64_t dim=0) {
+    return at::_ops::tensor_split_indices::call(self, c10::fromIntArrayRefSlow(indices), dim);
+  }
+}
+
+// aten::tensor_split.indices(Tensor(a -> *) self, SymInt[] indices, int dim=0) -> Tensor(a)[]
+inline ::std::vector<at::Tensor> tensor_split_symint(const at::Tensor & self, c10::SymIntArrayRef indices, int64_t dim=0) {
+    return at::_ops::tensor_split_indices::call(self, indices, dim);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  ::std::vector<at::Tensor> tensor_split(const at::Tensor & self, c10::SymIntArrayRef indices, int64_t dim=0) {
+    return at::_ops::tensor_split_indices::call(self, indices, dim);
+  }
+}
+
+// aten::tensor_split.tensor_indices_or_sections(Tensor(a -> *) self, Tensor tensor_indices_or_sections, int dim=0) -> Tensor(a)[]
+inline ::std::vector<at::Tensor> tensor_split(const at::Tensor & self, const at::Tensor & tensor_indices_or_sections, int64_t dim=0) {
+    return at::_ops::tensor_split_tensor_indices_or_sections::call(self, tensor_indices_or_sections, dim);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor_split_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor_split_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..267bfbc5f3ddf35709da8aeb8395760bd4cbb1f5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor_split_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::vector<at::Tensor> tensor_split_sections_symint(const at::Tensor & self, c10::SymInt sections, int64_t dim=0);
+TORCH_API ::std::vector<at::Tensor> tensor_split_indices_symint(const at::Tensor & self, c10::SymIntArrayRef indices, int64_t dim=0);
+TORCH_API ::std::vector<at::Tensor> tensor_split(const at::Tensor & self, const at::Tensor & tensor_indices_or_sections, int64_t dim=0);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor_split_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor_split_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..15f9079c6baef62fcb0a0f9caa3d1eb5d63eac60
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensor_split_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 tensor_split_sections {
+  using schema = ::std::vector<at::Tensor> (const at::Tensor &, c10::SymInt, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::tensor_split";
+  static constexpr const char* overload_name = "sections";
+  static constexpr const char* schema_str = "tensor_split.sections(Tensor(a -> *) self, SymInt sections, int dim=0) -> Tensor(a)[]";
+  static ::std::vector<at::Tensor> call(const at::Tensor & self, c10::SymInt sections, int64_t dim);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymInt sections, int64_t dim);
+};
+
+struct TORCH_API tensor_split_indices {
+  using schema = ::std::vector<at::Tensor> (const at::Tensor &, c10::SymIntArrayRef, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::tensor_split";
+  static constexpr const char* overload_name = "indices";
+  static constexpr const char* schema_str = "tensor_split.indices(Tensor(a -> *) self, SymInt[] indices, int dim=0) -> Tensor(a)[]";
+  static ::std::vector<at::Tensor> call(const at::Tensor & self, c10::SymIntArrayRef indices, int64_t dim);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef indices, int64_t dim);
+};
+
+struct TORCH_API tensor_split_tensor_indices_or_sections {
+  using schema = ::std::vector<at::Tensor> (const at::Tensor &, const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::tensor_split";
+  static constexpr const char* overload_name = "tensor_indices_or_sections";
+  static constexpr const char* schema_str = "tensor_split.tensor_indices_or_sections(Tensor(a -> *) self, Tensor tensor_indices_or_sections, int dim=0) -> Tensor(a)[]";
+  static ::std::vector<at::Tensor> call(const at::Tensor & self, const at::Tensor & tensor_indices_or_sections, int64_t dim);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & tensor_indices_or_sections, int64_t dim);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensordot.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensordot.h
new file mode 100644
index 0000000000000000000000000000000000000000..2482398abc04ca982bac29a5c3c49e23dd64ac0f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tensordot.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/tensordot_ops.h>
+
+namespace at {
+
+
+// aten::tensordot(Tensor self, Tensor other, int[] dims_self, int[] dims_other) -> Tensor
+inline at::Tensor tensordot(const at::Tensor & self, const at::Tensor & other, at::IntArrayRef dims_self, at::IntArrayRef dims_other) {
+    return at::_ops::tensordot::call(self, other, dims_self, dims_other);
+}
+
+// aten::tensordot.out(Tensor self, Tensor other, int[] dims_self, int[] dims_other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & tensordot_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other, at::IntArrayRef dims_self, at::IntArrayRef dims_other) {
+    return at::_ops::tensordot_out::call(self, other, dims_self, dims_other, out);
+}
+// aten::tensordot.out(Tensor self, Tensor other, int[] dims_self, int[] dims_other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & tensordot_outf(const at::Tensor & self, const at::Tensor & other, at::IntArrayRef dims_self, at::IntArrayRef dims_other, at::Tensor & out) {
+    return at::_ops::tensordot_out::call(self, other, dims_self, dims_other, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/threshold_backward_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/threshold_backward_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..ea8abc33ee79cf4fb8b6c93033d729928433c8d2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/threshold_backward_meta_dispatch.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 threshold_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & threshold);
+TORCH_API at::Tensor & threshold_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & threshold);
+TORCH_API at::Tensor & threshold_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & threshold, at::Tensor & grad_input);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/threshold_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/threshold_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..f1eccd3f56176414225e9f112d1dfb14f96ab9f2
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/threshold_native.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_threshold_out : public at::meta::structured_threshold {
+void impl(const at::Tensor & self, const at::Scalar & threshold, const at::Scalar & value, const at::Tensor & out);
+};
+TORCH_API at::Tensor threshold_quantized_cpu(const at::Tensor & self, const at::Scalar & threshold, const at::Scalar & value);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tile_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tile_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..2ef8f168df819976ed1c5f108ad65cf4bb999f61
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tile_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 tile {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::tile";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "tile(Tensor self, SymInt[] dims) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymIntArrayRef dims);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef dims);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to.h
new file mode 100644
index 0000000000000000000000000000000000000000..f4c93fcd04d9e2187e78798f49e7e0ce0dd7d408
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/to_ops.h>
+
+namespace at {
+
+
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_dense_backward.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_dense_backward.h
new file mode 100644
index 0000000000000000000000000000000000000000..232a65605721480ccedaed342d011c39b5818ec7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_dense_backward.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/to_dense_backward_ops.h>
+
+namespace at {
+
+
+// aten::to_dense_backward(Tensor grad, Tensor input, bool? masked_grad=None) -> Tensor
+inline at::Tensor to_dense_backward(const at::Tensor & grad, const at::Tensor & input, ::std::optional<bool> masked_grad=::std::nullopt) {
+    return at::_ops::to_dense_backward::call(grad, input, masked_grad);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_mkldnn_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_mkldnn_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..63e187e2de81e46852f31e16aed342475cd29f3b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_mkldnn_backward_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 const char* name = "aten::to_mkldnn_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* 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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_mkldnn_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_mkldnn_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..e924fdd3a297c0b7f9715df8a93dc710de39d15f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_mkldnn_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 to_mkldnn(const at::Tensor & self, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_sparse_bsc_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_sparse_bsc_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..f9b26a08b253ed009d8877c97a24123f0ede624a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_sparse_bsc_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 to_sparse_bsc(const at::Tensor & self, at::IntArrayRef blocksize, ::std::optional<int64_t> dense_dim=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_sparse_bsr_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_sparse_bsr_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..ba0242b5d160dd67cdd4c73049d4662b3520a150
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_sparse_bsr_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 to_sparse_bsr(const at::Tensor & self, at::IntArrayRef blocksize, ::std::optional<int64_t> dense_dim=::std::nullopt);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_sparse_csc_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_sparse_csc_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5d10329e298a1bed6d6f6d1d191051f879177843
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/to_sparse_csc_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 to_sparse_csc(const at::Tensor & self, ::std::optional<int64_t> dense_dim=::std::nullopt);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/topk_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/topk_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..b14ee15be3aeafd5b1cb5e8435f01d4ff304a4f4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/topk_native.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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/topk_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_topk_out_cpu : public at::meta::structured_topk {
+void impl(const at::Tensor & self, int64_t k, int64_t dim, bool largest, bool sorted, const at::Tensor & values, const at::Tensor & indices);
+};
+struct TORCH_API structured_topk_out_cuda : public at::meta::structured_topk {
+void impl(const at::Tensor & self, int64_t k, int64_t dim, bool largest, bool sorted, const at::Tensor & values, const at::Tensor & indices);
+};
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> topk_quantized_cpu(const at::Tensor & self, int64_t k, int64_t dim=-1, bool largest=true, bool sorted=true);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trace_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trace_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b0d4de39a37c2b4d413dccafffa6fef9376f7b24
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trace_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 trace(const at::Tensor & self);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/transpose_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/transpose_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b5b424a12bdfbd2d3c9f1ad52dca50aa0f25fe4e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/transpose_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 transpose(const at::Tensor & self, int64_t dim0, int64_t dim1);
+TORCH_API at::Tensor & transpose_(at::Tensor & self, int64_t dim0, int64_t dim1);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/transpose_copy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/transpose_copy.h
new file mode 100644
index 0000000000000000000000000000000000000000..eba0bc151c2b341173f6135abbf6f9020214c7eb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/transpose_copy.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/transpose_copy_ops.h>
+
+namespace at {
+
+
+// aten::transpose_copy.int(Tensor self, int dim0, int dim1) -> Tensor
+inline at::Tensor transpose_copy(const at::Tensor & self, int64_t dim0, int64_t dim1) {
+    return at::_ops::transpose_copy_int::call(self, dim0, dim1);
+}
+
+// aten::transpose_copy.int_out(Tensor self, int dim0, int dim1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & transpose_copy_out(at::Tensor & out, const at::Tensor & self, int64_t dim0, int64_t dim1) {
+    return at::_ops::transpose_copy_int_out::call(self, dim0, dim1, out);
+}
+// aten::transpose_copy.int_out(Tensor self, int dim0, int dim1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & transpose_copy_outf(const at::Tensor & self, int64_t dim0, int64_t dim1, at::Tensor & out) {
+    return at::_ops::transpose_copy_int_out::call(self, dim0, dim1, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/transpose_copy_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/transpose_copy_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..686da3adbe838aef91c903edcd07ec0e8af1facc
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/transpose_copy_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 transpose_copy(const at::Tensor & self, int64_t dim0, int64_t dim1);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trapz_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trapz_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..c7ddd54b73f1c801b4358190f37343ac0c0bafba
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trapz_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 trapz(const at::Tensor & y, const at::Tensor & x, int64_t dim=-1);
+TORCH_API at::Tensor trapz(const at::Tensor & y, double dx=1, int64_t dim=-1);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tril_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tril_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6e907ec0fe2a36f25e29df66bc683cd417ca3325
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tril_cpu_dispatch.h
@@ -0,0 +1,35 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 tril(const at::Tensor & self, int64_t diagonal=0);
+TORCH_API at::Tensor tril_symint(const at::Tensor & self, c10::SymInt diagonal=0);
+TORCH_API at::Tensor & tril_out(at::Tensor & out, const at::Tensor & self, int64_t diagonal=0);
+TORCH_API at::Tensor & tril_outf(const at::Tensor & self, int64_t diagonal, at::Tensor & out);
+TORCH_API at::Tensor & tril_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymInt diagonal=0);
+TORCH_API at::Tensor & tril_symint_outf(const at::Tensor & self, c10::SymInt diagonal, at::Tensor & out);
+TORCH_API at::Tensor & tril_(at::Tensor & self, int64_t diagonal=0);
+TORCH_API at::Tensor & tril__symint(at::Tensor & self, c10::SymInt diagonal=0);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tril_indices_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tril_indices_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..e628554f82f86d1271b9e23427cc715dd50765d0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/tril_indices_native.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & tril_indices_out(int64_t row, int64_t col, int64_t offset, at::Tensor & out);
+TORCH_API at::Tensor tril_indices_cpu(int64_t row, int64_t col, int64_t offset=0, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor tril_indices_cuda(int64_t row, int64_t col, int64_t offset=0, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/triplet_margin_loss.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/triplet_margin_loss.h
new file mode 100644
index 0000000000000000000000000000000000000000..aff3858f6338e5862209b15f5c56d158bbd07a5f
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/triplet_margin_loss.h
@@ -0,0 +1,36 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/triplet_margin_loss_ops.h>
+
+namespace at {
+
+
+// aten::triplet_margin_loss(Tensor anchor, Tensor positive, Tensor negative, float margin=1.0, float p=2, float eps=1e-06, bool swap=False, int reduction=Mean) -> Tensor
+inline at::Tensor triplet_margin_loss(const at::Tensor & anchor, const at::Tensor & positive, const at::Tensor & negative, double margin=1.0, double p=2, double eps=1e-06, bool swap=false, int64_t reduction=at::Reduction::Mean) {
+    return at::_ops::triplet_margin_loss::call(anchor, positive, negative, margin, p, eps, swap, reduction);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/true_divide_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/true_divide_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..07ce6e4e8a37cfa52a3959e151a37cf9cc873f7a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/true_divide_native.h
@@ -0,0 +1,30 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 true_divide(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & true_divide_out(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & true_divide_(at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor true_divide(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & true_divide_(at::Tensor & self, const at::Scalar & other);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trunc.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trunc.h
new file mode 100644
index 0000000000000000000000000000000000000000..1d5bf9ccb680ce05425bd5a6e8bfd973481319c7
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trunc.h
@@ -0,0 +1,50 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/trunc_ops.h>
+
+namespace at {
+
+
+// aten::trunc(Tensor self) -> Tensor
+inline at::Tensor trunc(const at::Tensor & self) {
+    return at::_ops::trunc::call(self);
+}
+
+// aten::trunc_(Tensor(a!) self) -> Tensor(a!)
+inline at::Tensor & trunc_(at::Tensor & self) {
+    return at::_ops::trunc_::call(self);
+}
+
+// aten::trunc.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & trunc_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::trunc_out::call(self, out);
+}
+// aten::trunc.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & trunc_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::trunc_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trunc_meta_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trunc_meta_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..6b20fa2645f3ad068e9a099094128371c279981d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trunc_meta_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 trunc(const at::Tensor & self);
+TORCH_API at::Tensor & trunc_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & trunc_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & trunc_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trunc_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trunc_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..0794d44241b1529c5f38b8f0337d2b8901a7ae95
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/trunc_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 trunc {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::trunc";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "trunc(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 trunc_ {
+  using schema = at::Tensor & (at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::trunc_";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "trunc_(Tensor(a!) self) -> Tensor(a!)";
+  static at::Tensor & call(at::Tensor & self);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self);
+};
+
+struct TORCH_API trunc_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 const char* name = "aten::trunc";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "trunc.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
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/type_as_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/type_as_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..8872a83f4bb621234265cccbe3bc51705bdd7fc1
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/type_as_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 type_as(const at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/type_as_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/type_as_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..b75388dc5a0a9a738a1cd6007cd74ff4d46785ec
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/type_as_ops.h
@@ -0,0 +1,34 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 type_as {
+  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 const char* name = "aten::type_as";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "type_as(Tensor self, Tensor other) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & other);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unflatten_dense_tensors_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unflatten_dense_tensors_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..a841e1b6031faaa573968d6b2075c97cf418f503
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unflatten_dense_tensors_native.h
@@ -0,0 +1,26 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::vector<at::Tensor> unflatten_dense_tensors(const at::Tensor & flat, at::TensorList tensors);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unflatten_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unflatten_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..784390e30335b224150250d6326c90171121eb43
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unflatten_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 unflatten_symint(const at::Tensor & self, int64_t dim, c10::SymIntArrayRef sizes);
+TORCH_API at::Tensor unflatten_dimname_symint(const at::Tensor & self, at::Dimname dim, c10::SymIntArrayRef sizes, at::DimnameList names);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unflatten_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unflatten_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..c8492037c22d847d2ed7aa94b6d59be9f18e7b8e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unflatten_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 unflatten_int {
+  using schema = at::Tensor (const at::Tensor &, int64_t, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::unflatten";
+  static constexpr const char* overload_name = "int";
+  static constexpr const char* schema_str = "unflatten.int(Tensor(a) self, int dim, SymInt[] sizes) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, int64_t dim, c10::SymIntArrayRef sizes);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, c10::SymIntArrayRef sizes);
+};
+
+struct TORCH_API unflatten_Dimname {
+  using schema = at::Tensor (const at::Tensor &, at::Dimname, c10::SymIntArrayRef, at::DimnameList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::unflatten";
+  static constexpr const char* overload_name = "Dimname";
+  static constexpr const char* schema_str = "unflatten.Dimname(Tensor(a) self, Dimname dim, SymInt[] sizes, Dimname[] names) -> Tensor(a)";
+  static at::Tensor call(const at::Tensor & self, at::Dimname dim, c10::SymIntArrayRef sizes, at::DimnameList names);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Dimname dim, c10::SymIntArrayRef sizes, at::DimnameList names);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unfold_backward_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unfold_backward_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5b3f492ea23c4a96edb3f08cf7cb32c0ab0c975a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unfold_backward_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & unfold_backward_out(at::Tensor & out, const at::Tensor & grad_in, at::IntArrayRef input_sizes, int64_t dim, int64_t size, int64_t step);
+TORCH_API at::Tensor & unfold_backward_outf(const at::Tensor & grad_in, at::IntArrayRef input_sizes, int64_t dim, int64_t size, int64_t step, at::Tensor & out);
+TORCH_API at::Tensor & unfold_backward_symint_out(at::Tensor & out, const at::Tensor & grad_in, c10::SymIntArrayRef input_sizes, int64_t dim, int64_t size, int64_t step);
+TORCH_API at::Tensor & unfold_backward_symint_outf(const at::Tensor & grad_in, c10::SymIntArrayRef input_sizes, int64_t dim, int64_t size, int64_t step, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unfold_backward_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unfold_backward_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d0379fcfa0d812120afa20bbd52453af91ac58e4
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unfold_backward_cpu_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 unfold_backward(const at::Tensor & grad_in, at::IntArrayRef input_sizes, int64_t dim, int64_t size, int64_t step);
+TORCH_API at::Tensor unfold_backward_symint(const at::Tensor & grad_in, c10::SymIntArrayRef input_sizes, int64_t dim, int64_t size, int64_t step);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unfold_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unfold_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d55e9091ae9d6b4c6860a7c451e8a4131b51f1d5
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unfold_cpu_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 unfold(const at::Tensor & self, int64_t dimension, int64_t size, int64_t step);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unique_consecutive.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unique_consecutive.h
new file mode 100644
index 0000000000000000000000000000000000000000..4aa712912a00128867cceedc789833c0dda2e747
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unique_consecutive.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/unique_consecutive_ops.h>
+
+namespace at {
+
+
+// aten::unique_consecutive(Tensor self, bool return_inverse=False, bool return_counts=False, int? dim=None) -> (Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor> unique_consecutive(const at::Tensor & self, bool return_inverse=false, bool return_counts=false, ::std::optional<int64_t> dim=::std::nullopt) {
+    return at::_ops::unique_consecutive::call(self, return_inverse, return_counts, dim);
+}
+
+// aten::unique_consecutive.out(Tensor self, bool return_inverse=False, bool return_counts=False, int? dim=None, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> unique_consecutive_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & self, bool return_inverse=false, bool return_counts=false, ::std::optional<int64_t> dim=::std::nullopt) {
+    return at::_ops::unique_consecutive_out::call(self, return_inverse, return_counts, dim, out0, out1, out2);
+}
+// aten::unique_consecutive.out(Tensor self, bool return_inverse=False, bool return_counts=False, int? dim=None, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> unique_consecutive_outf(const at::Tensor & self, bool return_inverse, bool return_counts, ::std::optional<int64_t> dim, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2) {
+    return at::_ops::unique_consecutive_out::call(self, return_inverse, return_counts, dim, out0, out1, out2);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unique_dim_consecutive.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unique_dim_consecutive.h
new file mode 100644
index 0000000000000000000000000000000000000000..57eee21ea88e999bc01b3731174e5d7567bcaf5b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/unique_dim_consecutive.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/unique_dim_consecutive_ops.h>
+
+namespace at {
+
+
+// aten::unique_dim_consecutive(Tensor self, int dim, bool return_inverse=False, bool return_counts=False) -> (Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor> unique_dim_consecutive(const at::Tensor & self, int64_t dim, bool return_inverse=false, bool return_counts=false) {
+    return at::_ops::unique_dim_consecutive::call(self, dim, return_inverse, return_counts);
+}
+
+// aten::unique_dim_consecutive.out(Tensor self, int dim, bool return_inverse=False, bool return_counts=False, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> unique_dim_consecutive_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & self, int64_t dim, bool return_inverse=false, bool return_counts=false) {
+    return at::_ops::unique_dim_consecutive_out::call(self, dim, return_inverse, return_counts, out0, out1, out2);
+}
+// aten::unique_dim_consecutive.out(Tensor self, int dim, bool return_inverse=False, bool return_counts=False, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> unique_dim_consecutive_outf(const at::Tensor & self, int64_t dim, bool return_inverse, bool return_counts, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2) {
+    return at::_ops::unique_dim_consecutive_out::call(self, dim, return_inverse, return_counts, out0, out1, out2);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d.h
new file mode 100644
index 0000000000000000000000000000000000000000..40ac94660d1123313c0918aa5f985dae27c4eaae
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d.h
@@ -0,0 +1,119 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/upsample_bicubic2d_ops.h>
+
+namespace at {
+
+
+// aten::upsample_bicubic2d.vec(Tensor input, SymInt[]? output_size, bool align_corners, float[]? scale_factors) -> Tensor
+inline at::Tensor upsample_bicubic2d(const at::Tensor & input, at::OptionalIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bicubic2d_vec::call(input, output_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*output_size)) : ::std::nullopt, align_corners, scale_factors);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor upsample_bicubic2d(const at::Tensor & input, at::OptionalIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bicubic2d_vec::call(input, output_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*output_size)) : ::std::nullopt, align_corners, scale_factors);
+  }
+}
+
+// aten::upsample_bicubic2d.vec(Tensor input, SymInt[]? output_size, bool align_corners, float[]? scale_factors) -> Tensor
+inline at::Tensor upsample_bicubic2d_symint(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bicubic2d_vec::call(input, output_size, align_corners, scale_factors);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor upsample_bicubic2d(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bicubic2d_vec::call(input, output_size, align_corners, scale_factors);
+  }
+}
+
+// aten::upsample_bicubic2d.out(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bicubic2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_out::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & upsample_bicubic2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_out::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w, out);
+  }
+}
+
+// aten::upsample_bicubic2d.out(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bicubic2d_outf(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::upsample_bicubic2d_out::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & upsample_bicubic2d_outf(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::upsample_bicubic2d_out::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w, out);
+  }
+}
+
+// aten::upsample_bicubic2d.out(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bicubic2d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_out::call(self, output_size, align_corners, scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & upsample_bicubic2d_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d_out::call(self, output_size, align_corners, scales_h, scales_w, out);
+  }
+}
+
+// aten::upsample_bicubic2d.out(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bicubic2d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::upsample_bicubic2d_out::call(self, output_size, align_corners, scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & upsample_bicubic2d_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::upsample_bicubic2d_out::call(self, output_size, align_corners, scales_h, scales_w, out);
+  }
+}
+
+// aten::upsample_bicubic2d(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor
+inline at::Tensor upsample_bicubic2d(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor upsample_bicubic2d(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w);
+  }
+}
+
+// aten::upsample_bicubic2d(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor
+inline at::Tensor upsample_bicubic2d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d::call(self, output_size, align_corners, scales_h, scales_w);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor upsample_bicubic2d(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bicubic2d::call(self, output_size, align_corners, scales_h, scales_w);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_backward_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_backward_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..173fc82eebb383b7fbc07bd2f1ca11cb4b1eee44
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_backward_cuda_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_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);
+TORCH_API 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);
+TORCH_API 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);
+TORCH_API 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);
+TORCH_API 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);
+TORCH_API 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);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3c1a3374e93b41491d0cb65acb35406f31c0c878
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_bicubic2d(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor upsample_bicubic2d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_cpu_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_cpu_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..c260c16ac44aef7cfe42b886d8c24af67e868959
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_cpu_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_bicubic2d(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor upsample_bicubic2d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & upsample_bicubic2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & upsample_bicubic2d_outf(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+TORCH_API at::Tensor & upsample_bicubic2d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & upsample_bicubic2d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..5a3d7de31f90b7e867cdc1f09a5867aa68419625
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_cuda_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_bicubic2d(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor upsample_bicubic2d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & upsample_bicubic2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & upsample_bicubic2d_outf(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+TORCH_API at::Tensor & upsample_bicubic2d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt);
+TORCH_API at::Tensor & upsample_bicubic2d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..9dcb9d4b6a016b9b393e4f88a6d7b6a82100bbb9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bicubic2d_ops.h
@@ -0,0 +1,56 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API upsample_bicubic2d_vec {
+  using schema = at::Tensor (const at::Tensor &, at::OptionalSymIntArrayRef, bool, ::std::optional<at::ArrayRef<double>>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::upsample_bicubic2d";
+  static constexpr const char* overload_name = "vec";
+  static constexpr const char* schema_str = "upsample_bicubic2d.vec(Tensor input, SymInt[]? output_size, bool align_corners, float[]? scale_factors) -> Tensor";
+  static at::Tensor call(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, at::OptionalSymIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors);
+};
+
+struct TORCH_API upsample_bicubic2d_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymIntArrayRef, bool, ::std::optional<double>, ::std::optional<double>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::upsample_bicubic2d";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "upsample_bicubic2d.out(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+};
+
+struct TORCH_API upsample_bicubic2d {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef, bool, ::std::optional<double>, ::std::optional<double>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::upsample_bicubic2d";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "upsample_bicubic2d(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor";
+  static at::Tensor call(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bilinear2d.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bilinear2d.h
new file mode 100644
index 0000000000000000000000000000000000000000..8225001fee5d5a8b87d438a1ae1884e2d5a26edf
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bilinear2d.h
@@ -0,0 +1,163 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/upsample_bilinear2d_ops.h>
+
+namespace at {
+
+
+// aten::upsample_bilinear2d.vec(Tensor input, SymInt[]? output_size, bool align_corners, float[]? scale_factors) -> Tensor
+inline at::Tensor upsample_bilinear2d(const at::Tensor & input, at::OptionalIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bilinear2d_vec::call(input, output_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*output_size)) : ::std::nullopt, align_corners, scale_factors);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor upsample_bilinear2d(const at::Tensor & input, at::OptionalIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bilinear2d_vec::call(input, output_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*output_size)) : ::std::nullopt, align_corners, scale_factors);
+  }
+}
+
+// aten::upsample_bilinear2d.vec(Tensor input, SymInt[]? output_size, bool align_corners, float[]? scale_factors) -> Tensor
+inline at::Tensor upsample_bilinear2d_symint(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bilinear2d_vec::call(input, output_size, align_corners, scale_factors);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor upsample_bilinear2d(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bilinear2d_vec::call(input, output_size, align_corners, scale_factors);
+  }
+}
+
+// aten::upsample_bilinear2d.out(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bilinear2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bilinear2d_out::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & upsample_bilinear2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bilinear2d_out::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w, out);
+  }
+}
+
+// aten::upsample_bilinear2d.out(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bilinear2d_outf(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::upsample_bilinear2d_out::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & upsample_bilinear2d_outf(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::upsample_bilinear2d_out::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w, out);
+  }
+}
+
+// aten::upsample_bilinear2d.out(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bilinear2d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bilinear2d_out::call(self, output_size, align_corners, scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & upsample_bilinear2d_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bilinear2d_out::call(self, output_size, align_corners, scales_h, scales_w, out);
+  }
+}
+
+// aten::upsample_bilinear2d.out(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bilinear2d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::upsample_bilinear2d_out::call(self, output_size, align_corners, scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & upsample_bilinear2d_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::upsample_bilinear2d_out::call(self, output_size, align_corners, scales_h, scales_w, out);
+  }
+}
+
+// aten::upsample_bilinear2d(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor
+inline at::Tensor upsample_bilinear2d(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bilinear2d::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor upsample_bilinear2d(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bilinear2d::call(self, c10::fromIntArrayRefSlow(output_size), align_corners, scales_h, scales_w);
+  }
+}
+
+// aten::upsample_bilinear2d(Tensor self, SymInt[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor
+inline at::Tensor upsample_bilinear2d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bilinear2d::call(self, output_size, align_corners, scales_h, scales_w);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor upsample_bilinear2d(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_h=::std::nullopt, ::std::optional<double> scales_w=::std::nullopt) {
+    return at::_ops::upsample_bilinear2d::call(self, output_size, align_corners, scales_h, scales_w);
+  }
+}
+
+// aten::upsample_bilinear2d.vec_out(Tensor input, SymInt[]? output_size, bool align_corners, float[]? scale_factors, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bilinear2d_out(at::Tensor & out, const at::Tensor & input, at::OptionalIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bilinear2d_vec_out::call(input, output_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*output_size)) : ::std::nullopt, align_corners, scale_factors, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & upsample_bilinear2d_out(at::Tensor & out, const at::Tensor & input, at::OptionalIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bilinear2d_vec_out::call(input, output_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*output_size)) : ::std::nullopt, align_corners, scale_factors, out);
+  }
+}
+
+// aten::upsample_bilinear2d.vec_out(Tensor input, SymInt[]? output_size, bool align_corners, float[]? scale_factors, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bilinear2d_outf(const at::Tensor & input, at::OptionalIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors, at::Tensor & out) {
+    return at::_ops::upsample_bilinear2d_vec_out::call(input, output_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*output_size)) : ::std::nullopt, align_corners, scale_factors, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & upsample_bilinear2d_outf(const at::Tensor & input, at::OptionalIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors, at::Tensor & out) {
+    return at::_ops::upsample_bilinear2d_vec_out::call(input, output_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*output_size)) : ::std::nullopt, align_corners, scale_factors, out);
+  }
+}
+
+// aten::upsample_bilinear2d.vec_out(Tensor input, SymInt[]? output_size, bool align_corners, float[]? scale_factors, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bilinear2d_symint_out(at::Tensor & out, const at::Tensor & input, at::OptionalSymIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bilinear2d_vec_out::call(input, output_size, align_corners, scale_factors, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & upsample_bilinear2d_out(at::Tensor & out, const at::Tensor & input, at::OptionalSymIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::upsample_bilinear2d_vec_out::call(input, output_size, align_corners, scale_factors, out);
+  }
+}
+
+// aten::upsample_bilinear2d.vec_out(Tensor input, SymInt[]? output_size, bool align_corners, float[]? scale_factors, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & upsample_bilinear2d_symint_outf(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors, at::Tensor & out) {
+    return at::_ops::upsample_bilinear2d_vec_out::call(input, output_size, align_corners, scale_factors, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & upsample_bilinear2d_outf(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors, at::Tensor & out) {
+    return at::_ops::upsample_bilinear2d_vec_out::call(input, output_size, align_corners, scale_factors, out);
+  }
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bilinear2d_backward_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bilinear2d_backward_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..19df49aa5e2d052fdbd28cde1f86f78131901f19
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_bilinear2d_backward_ops.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API upsample_bilinear2d_backward_grad_input {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, ::std::optional<double>, ::std::optional<double>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::upsample_bilinear2d_backward";
+  static constexpr const char* overload_name = "grad_input";
+  static constexpr const char* schema_str = "upsample_bilinear2d_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!)";
+  static at::Tensor & call(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);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, 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);
+};
+
+struct TORCH_API upsample_bilinear2d_backward {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, ::std::optional<double>, ::std::optional<double>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::upsample_bilinear2d_backward";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "upsample_bilinear2d_backward(Tensor grad_output, SymInt[2] output_size, SymInt[4] input_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor";
+  static at::Tensor call(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);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, 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);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_linear1d_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_linear1d_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9298cdc9d0e886598b39370bd543f1036efaf8a8
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_linear1d_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_linear1d(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales=::std::nullopt);
+TORCH_API at::Tensor upsample_linear1d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales=::std::nullopt);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_nearest1d_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_nearest1d_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f254550920e3034430d37db3791c35d6ab660d88
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_nearest1d_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 upsample_nearest1d(const at::Tensor & input, at::OptionalIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors);
+TORCH_API at::Tensor upsample_nearest1d_symint(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_nearest2d_backward_meta.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_nearest2d_backward_meta.h
new file mode 100644
index 0000000000000000000000000000000000000000..74e473789eee2e45465cb0b88b3627005d9a860a
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_nearest2d_backward_meta.h
@@ -0,0 +1,32 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_upsample_nearest2d_backward : public at::impl::MetaBase {
+
+
+    void meta(const at::Tensor & grad_output, at::ArrayRef<int64_t> output_size, at::ArrayRef<int64_t> input_size, ::std::optional<double> scales_h, ::std::optional<double> scales_w);
+};
+
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_nearest2d_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_nearest2d_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..3e5d2cb4b2a174021cfa41cf2441f6fa6cd4c351
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_nearest2d_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 & upsample_nearest2d_out(at::Tensor & out, const at::Tensor & input, at::OptionalIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors);
+TORCH_API at::Tensor & upsample_nearest2d_outf(const at::Tensor & input, at::OptionalIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors, at::Tensor & out);
+TORCH_API at::Tensor & upsample_nearest2d_symint_out(at::Tensor & out, const at::Tensor & input, at::OptionalSymIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors);
+TORCH_API at::Tensor & upsample_nearest2d_symint_outf(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_trilinear3d_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_trilinear3d_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..883f0cfa96f21d73ee425f1f14a0fe459058e338
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/upsample_trilinear3d_cuda_dispatch.h
@@ -0,0 +1,33 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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_trilinear3d(const at::Tensor & self, at::IntArrayRef output_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_symint(const at::Tensor & self, c10::SymIntArrayRef output_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_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_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_outf(const at::Tensor & self, at::IntArrayRef output_size, bool align_corners, ::std::optional<double> scales_d, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+TORCH_API at::Tensor & upsample_trilinear3d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_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_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, bool align_corners, ::std::optional<double> scales_d, ::std::optional<double> scales_h, ::std::optional<double> scales_w, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/value_selecting_reduction_backward_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/value_selecting_reduction_backward_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..d36bf74f1321f1af544716ff916eca3f79902a9c
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/value_selecting_reduction_backward_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 value_selecting_reduction_backward(const at::Tensor & grad, int64_t dim, const at::Tensor & indices, at::IntArrayRef sizes, bool keepdim);
+TORCH_API at::Tensor value_selecting_reduction_backward_symint(const at::Tensor & grad, int64_t dim, const at::Tensor & indices, c10::SymIntArrayRef sizes, bool keepdim);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/values_copy_compositeexplicitautogradnonfunctional_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/values_copy_compositeexplicitautogradnonfunctional_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..f740afcf0b25217e2510411eca318b0b8395bf74
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/values_copy_compositeexplicitautogradnonfunctional_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 values_copy(const at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/var_mean_compositeexplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/var_mean_compositeexplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..b3295d1a19bc7b5a1c8077e767ff5ea18137d37b
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/var_mean_compositeexplicitautograd_dispatch.h
@@ -0,0 +1,29 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 &> var_mean_out(at::Tensor & out0, at::Tensor & out1, const at::Tensor & self, at::OptionalIntArrayRef dim=::std::nullopt, const ::std::optional<at::Scalar> & correction=::std::nullopt, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> var_mean_outf(const at::Tensor & self, at::OptionalIntArrayRef dim, const ::std::optional<at::Scalar> & correction, bool keepdim, at::Tensor & out0, at::Tensor & out1);
+
+} // namespace compositeexplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/var_mean_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/var_mean_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..284326515f8446fb937b9950c307562e85389629
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/var_mean_ops.h
@@ -0,0 +1,89 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 var_mean {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::var_mean";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "var_mean(Tensor self, bool unbiased=True) -> (Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, bool unbiased);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, bool unbiased);
+};
+
+struct TORCH_API var_mean_dim {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, at::OptionalIntArrayRef, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::var_mean";
+  static constexpr const char* overload_name = "dim";
+  static constexpr const char* schema_str = "var_mean.dim(Tensor self, int[1]? dim, bool unbiased=True, bool keepdim=False) -> (Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, at::OptionalIntArrayRef dim, bool unbiased, bool keepdim);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::OptionalIntArrayRef dim, bool unbiased, bool keepdim);
+};
+
+struct TORCH_API var_mean_correction {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, at::OptionalIntArrayRef, const ::std::optional<at::Scalar> &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::var_mean";
+  static constexpr const char* overload_name = "correction";
+  static constexpr const char* schema_str = "var_mean.correction(Tensor self, int[1]? dim=None, *, Scalar? correction=None, bool keepdim=False) -> (Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, at::OptionalIntArrayRef dim, const ::std::optional<at::Scalar> & correction, bool keepdim);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::OptionalIntArrayRef dim, const ::std::optional<at::Scalar> & correction, bool keepdim);
+};
+
+struct TORCH_API var_mean_names_dim {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, at::DimnameList, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::var_mean";
+  static constexpr const char* overload_name = "names_dim";
+  static constexpr const char* schema_str = "var_mean.names_dim(Tensor self, Dimname[1] dim, bool unbiased=True, bool keepdim=False) -> (Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, at::DimnameList dim, bool unbiased, bool keepdim);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::DimnameList dim, bool unbiased, bool keepdim);
+};
+
+struct TORCH_API var_mean_correction_names {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, at::DimnameList, const ::std::optional<at::Scalar> &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::var_mean";
+  static constexpr const char* overload_name = "correction_names";
+  static constexpr const char* schema_str = "var_mean.correction_names(Tensor self, Dimname[1] dim, *, Scalar? correction=None, bool keepdim=False) -> (Tensor, Tensor)";
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, at::DimnameList dim, const ::std::optional<at::Scalar> & correction, bool keepdim);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::DimnameList dim, const ::std::optional<at::Scalar> & correction, bool keepdim);
+};
+
+struct TORCH_API var_mean_correction_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, at::OptionalIntArrayRef, const ::std::optional<at::Scalar> &, bool, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::var_mean";
+  static constexpr const char* overload_name = "correction_out";
+  static constexpr const char* schema_str = "var_mean.correction_out(Tensor self, int[1]? dim=None, *, Scalar? correction=None, bool keepdim=False, Tensor(a!) out0, Tensor(b!) out1) -> (Tensor(a!), Tensor(b!))";
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, at::OptionalIntArrayRef dim, const ::std::optional<at::Scalar> & correction, bool keepdim, at::Tensor & out0, at::Tensor & out1);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::OptionalIntArrayRef dim, const ::std::optional<at::Scalar> & correction, bool keepdim, at::Tensor & out0, at::Tensor & out1);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/vdot_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/vdot_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..4cf92f24d63d1ead05331a57b7aac6f74165a40d
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/vdot_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 vdot(const at::Tensor & self, const at::Tensor & other);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/view_as_complex_cuda_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/view_as_complex_cuda_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..7460df92c779dd9c7d1869a684ee495593cf80ac
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/view_as_complex_cuda_dispatch.h
@@ -0,0 +1,28 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 view_as_complex(const at::Tensor & self);
+
+} // namespace cuda
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/view_as_real_copy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/view_as_real_copy.h
new file mode 100644
index 0000000000000000000000000000000000000000..7622dbff775464e82728560b143a3387851e7431
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/view_as_real_copy.h
@@ -0,0 +1,45 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/view_as_real_copy_ops.h>
+
+namespace at {
+
+
+// aten::view_as_real_copy(Tensor self) -> Tensor
+inline at::Tensor view_as_real_copy(const at::Tensor & self) {
+    return at::_ops::view_as_real_copy::call(self);
+}
+
+// aten::view_as_real_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & view_as_real_copy_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::view_as_real_copy_out::call(self, out);
+}
+// aten::view_as_real_copy.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & view_as_real_copy_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::view_as_real_copy_out::call(self, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/view_copy.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/view_copy.h
new file mode 100644
index 0000000000000000000000000000000000000000..6c5bae444bc61bc7d9f2ff505feede9256d76d88
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/view_copy.h
@@ -0,0 +1,111 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/view_copy_ops.h>
+
+namespace at {
+
+
+// aten::view_copy(Tensor self, SymInt[] size) -> Tensor
+inline at::Tensor view_copy(const at::Tensor & self, at::IntArrayRef size) {
+    return at::_ops::view_copy::call(self, c10::fromIntArrayRefSlow(size));
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor view_copy(const at::Tensor & self, at::IntArrayRef size) {
+    return at::_ops::view_copy::call(self, c10::fromIntArrayRefSlow(size));
+  }
+}
+
+// aten::view_copy(Tensor self, SymInt[] size) -> Tensor
+inline at::Tensor view_copy_symint(const at::Tensor & self, c10::SymIntArrayRef size) {
+    return at::_ops::view_copy::call(self, size);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor view_copy(const at::Tensor & self, c10::SymIntArrayRef size) {
+    return at::_ops::view_copy::call(self, size);
+  }
+}
+
+// aten::view_copy.dtype(Tensor self, ScalarType dtype) -> Tensor
+inline at::Tensor view_copy(const at::Tensor & self, at::ScalarType dtype) {
+    return at::_ops::view_copy_dtype::call(self, dtype);
+}
+
+// aten::view_copy.out(Tensor self, SymInt[] size, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & view_copy_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef size) {
+    return at::_ops::view_copy_out::call(self, c10::fromIntArrayRefSlow(size), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & view_copy_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef size) {
+    return at::_ops::view_copy_out::call(self, c10::fromIntArrayRefSlow(size), out);
+  }
+}
+
+// aten::view_copy.out(Tensor self, SymInt[] size, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & view_copy_outf(const at::Tensor & self, at::IntArrayRef size, at::Tensor & out) {
+    return at::_ops::view_copy_out::call(self, c10::fromIntArrayRefSlow(size), out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, int64_t>>>
+  at::Tensor & view_copy_outf(const at::Tensor & self, at::IntArrayRef size, at::Tensor & out) {
+    return at::_ops::view_copy_out::call(self, c10::fromIntArrayRefSlow(size), out);
+  }
+}
+
+// aten::view_copy.out(Tensor self, SymInt[] size, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & view_copy_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef size) {
+    return at::_ops::view_copy_out::call(self, size, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & view_copy_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef size) {
+    return at::_ops::view_copy_out::call(self, size, out);
+  }
+}
+
+// aten::view_copy.out(Tensor self, SymInt[] size, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & view_copy_symint_outf(const at::Tensor & self, c10::SymIntArrayRef size, at::Tensor & out) {
+    return at::_ops::view_copy_out::call(self, size, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same_v<T, c10::SymInt>>>
+  at::Tensor & view_copy_outf(const at::Tensor & self, c10::SymIntArrayRef size, at::Tensor & out) {
+    return at::_ops::view_copy_out::call(self, size, out);
+  }
+}
+
+// aten::view_copy.dtype_out(Tensor self, ScalarType dtype, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & view_copy_out(at::Tensor & out, const at::Tensor & self, at::ScalarType dtype) {
+    return at::_ops::view_copy_dtype_out::call(self, dtype, out);
+}
+// aten::view_copy.dtype_out(Tensor self, ScalarType dtype, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & view_copy_outf(const at::Tensor & self, at::ScalarType dtype, at::Tensor & out) {
+    return at::_ops::view_copy_dtype_out::call(self, dtype, out);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/vsplit_native.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/vsplit_native.h
new file mode 100644
index 0000000000000000000000000000000000000000..721292f36174bf5877e42ec48559c1acbbc9df8e
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/vsplit_native.h
@@ -0,0 +1,27 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::vector<at::Tensor> vsplit(const at::Tensor & self, int64_t sections);
+TORCH_API ::std::vector<at::Tensor> vsplit(const at::Tensor & self, at::IntArrayRef indices);
+} // namespace native
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/xor.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/xor.h
new file mode 100644
index 0000000000000000000000000000000000000000..80dc5b3ba7e2906b77b74ecd03cd46db6dc894cb
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/xor.h
@@ -0,0 +1,41 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 <string_view>
+
+
+
+#include <ATen/ops/xor_ops.h>
+
+namespace at {
+
+
+// aten::__xor__.Scalar(Tensor self, Scalar other) -> Tensor
+inline at::Tensor __xor__(const at::Tensor & self, const at::Scalar & other) {
+    return at::_ops::__xor___Scalar::call(self, other);
+}
+
+// aten::__xor__.Tensor(Tensor self, Tensor other) -> Tensor
+inline at::Tensor __xor__(const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::__xor___Tensor::call(self, other);
+}
+
+}
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/xor_compositeimplicitautograd_dispatch.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/xor_compositeimplicitautograd_dispatch.h
new file mode 100644
index 0000000000000000000000000000000000000000..9a70f0c795309fe590f14d6e2ebd5faf47adc1ab
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/xor_compositeimplicitautograd_dispatch.h
@@ -0,0 +1,31 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#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 __xor__(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & __ixor__(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor __xor__(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & __ixor__(at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeimplicitautograd
+} // namespace at
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
diff --git a/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/zeros_ops.h b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/zeros_ops.h
new file mode 100644
index 0000000000000000000000000000000000000000..b470cd5774fe486a96e69309bdb1e1821ce78120
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/torch/include/ATen/ops/zeros_ops.h
@@ -0,0 +1,67 @@
+#if !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <string_view>
+#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 zeros_names {
+  using schema = at::Tensor (at::IntArrayRef, ::std::optional<at::DimnameList>, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::zeros";
+  static constexpr const char* overload_name = "names";
+  static constexpr const char* schema_str = "zeros.names(int[] size, *, Dimname[]? names, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor";
+  static at::Tensor call(at::IntArrayRef size, ::std::optional<at::DimnameList> names, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, at::IntArrayRef size, ::std::optional<at::DimnameList> names, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+struct TORCH_API zeros {
+  using schema = at::Tensor (c10::SymIntArrayRef, ::std::optional<at::ScalarType>, ::std::optional<at::Layout>, ::std::optional<at::Device>, ::std::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::zeros";
+  static constexpr const char* overload_name = "";
+  static constexpr const char* schema_str = "zeros(SymInt[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor";
+  static at::Tensor call(c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, c10::SymIntArrayRef size, ::std::optional<at::ScalarType> dtype, ::std::optional<at::Layout> layout, ::std::optional<at::Device> device, ::std::optional<bool> pin_memory);
+};
+
+struct TORCH_API zeros_out {
+  using schema = at::Tensor & (c10::SymIntArrayRef, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::zeros";
+  static constexpr const char* overload_name = "out";
+  static constexpr const char* schema_str = "zeros.out(SymInt[] size, *, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(c10::SymIntArrayRef size, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, c10::SymIntArrayRef size, at::Tensor & out);
+};
+
+struct TORCH_API zeros_names_out {
+  using schema = at::Tensor & (at::IntArrayRef, ::std::optional<at::DimnameList>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  static constexpr const char* name = "aten::zeros";
+  static constexpr const char* overload_name = "names_out";
+  static constexpr const char* schema_str = "zeros.names_out(int[] size, *, Dimname[]? names, Tensor(a!) out) -> Tensor(a!)";
+  static at::Tensor & call(at::IntArrayRef size, ::std::optional<at::DimnameList> names, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::IntArrayRef size, ::std::optional<at::DimnameList> names, at::Tensor & out);
+};
+
+}} // namespace at::_ops
+
+#else
+#error "This file should not be included when either TORCH_STABLE_ONLY or TORCH_TARGET_VERSION is defined."
+#endif  // !defined(TORCH_STABLE_ONLY) && !defined(TORCH_TARGET_VERSION)