Add files using upload-large-folder tool

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

@@ -0,0 +1,91 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/_cudnn_rnn_flatten_weight_ops.h>
+
+namespace at {
+
+
+// aten::_cudnn_rnn_flatten_weight(Tensor[] weight_arr, int weight_stride0, SymInt input_size, int mode, SymInt hidden_size, SymInt proj_size, int num_layers, bool batch_first, bool bidirectional) -> Tensor
+inline at::Tensor _cudnn_rnn_flatten_weight(at::TensorList weight_arr, int64_t weight_stride0, int64_t input_size, int64_t mode, int64_t hidden_size, int64_t proj_size, int64_t num_layers, bool batch_first, bool bidirectional) {
+    return at::_ops::_cudnn_rnn_flatten_weight::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor _cudnn_rnn_flatten_weight(at::TensorList weight_arr, int64_t weight_stride0, int64_t input_size, int64_t mode, int64_t hidden_size, int64_t proj_size, int64_t num_layers, bool batch_first, bool bidirectional) {
+    return at::_ops::_cudnn_rnn_flatten_weight::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional);
+  }
+}
+
+// aten::_cudnn_rnn_flatten_weight(Tensor[] weight_arr, int weight_stride0, SymInt input_size, int mode, SymInt hidden_size, SymInt proj_size, int num_layers, bool batch_first, bool bidirectional) -> Tensor
+inline at::Tensor _cudnn_rnn_flatten_weight_symint(at::TensorList weight_arr, int64_t weight_stride0, c10::SymInt input_size, int64_t mode, c10::SymInt hidden_size, c10::SymInt proj_size, int64_t num_layers, bool batch_first, bool bidirectional) {
+    return at::_ops::_cudnn_rnn_flatten_weight::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor _cudnn_rnn_flatten_weight(at::TensorList weight_arr, int64_t weight_stride0, c10::SymInt input_size, int64_t mode, c10::SymInt hidden_size, c10::SymInt proj_size, int64_t num_layers, bool batch_first, bool bidirectional) {
+    return at::_ops::_cudnn_rnn_flatten_weight::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional);
+  }
+}
+
+// aten::_cudnn_rnn_flatten_weight.out(Tensor[] weight_arr, int weight_stride0, SymInt input_size, int mode, SymInt hidden_size, SymInt proj_size, int num_layers, bool batch_first, bool bidirectional, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _cudnn_rnn_flatten_weight_out(at::Tensor & out, at::TensorList weight_arr, int64_t weight_stride0, int64_t input_size, int64_t mode, int64_t hidden_size, int64_t proj_size, int64_t num_layers, bool batch_first, bool bidirectional) {
+    return at::_ops::_cudnn_rnn_flatten_weight_out::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & _cudnn_rnn_flatten_weight_out(at::Tensor & out, at::TensorList weight_arr, int64_t weight_stride0, int64_t input_size, int64_t mode, int64_t hidden_size, int64_t proj_size, int64_t num_layers, bool batch_first, bool bidirectional) {
+    return at::_ops::_cudnn_rnn_flatten_weight_out::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional, out);
+  }
+}
+
+// aten::_cudnn_rnn_flatten_weight.out(Tensor[] weight_arr, int weight_stride0, SymInt input_size, int mode, SymInt hidden_size, SymInt proj_size, int num_layers, bool batch_first, bool bidirectional, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _cudnn_rnn_flatten_weight_outf(at::TensorList weight_arr, int64_t weight_stride0, int64_t input_size, int64_t mode, int64_t hidden_size, int64_t proj_size, int64_t num_layers, bool batch_first, bool bidirectional, at::Tensor & out) {
+    return at::_ops::_cudnn_rnn_flatten_weight_out::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & _cudnn_rnn_flatten_weight_outf(at::TensorList weight_arr, int64_t weight_stride0, int64_t input_size, int64_t mode, int64_t hidden_size, int64_t proj_size, int64_t num_layers, bool batch_first, bool bidirectional, at::Tensor & out) {
+    return at::_ops::_cudnn_rnn_flatten_weight_out::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional, out);
+  }
+}
+
+// aten::_cudnn_rnn_flatten_weight.out(Tensor[] weight_arr, int weight_stride0, SymInt input_size, int mode, SymInt hidden_size, SymInt proj_size, int num_layers, bool batch_first, bool bidirectional, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _cudnn_rnn_flatten_weight_symint_out(at::Tensor & out, at::TensorList weight_arr, int64_t weight_stride0, c10::SymInt input_size, int64_t mode, c10::SymInt hidden_size, c10::SymInt proj_size, int64_t num_layers, bool batch_first, bool bidirectional) {
+    return at::_ops::_cudnn_rnn_flatten_weight_out::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & _cudnn_rnn_flatten_weight_out(at::Tensor & out, at::TensorList weight_arr, int64_t weight_stride0, c10::SymInt input_size, int64_t mode, c10::SymInt hidden_size, c10::SymInt proj_size, int64_t num_layers, bool batch_first, bool bidirectional) {
+    return at::_ops::_cudnn_rnn_flatten_weight_out::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional, out);
+  }
+}
+
+// aten::_cudnn_rnn_flatten_weight.out(Tensor[] weight_arr, int weight_stride0, SymInt input_size, int mode, SymInt hidden_size, SymInt proj_size, int num_layers, bool batch_first, bool bidirectional, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _cudnn_rnn_flatten_weight_symint_outf(at::TensorList weight_arr, int64_t weight_stride0, c10::SymInt input_size, int64_t mode, c10::SymInt hidden_size, c10::SymInt proj_size, int64_t num_layers, bool batch_first, bool bidirectional, at::Tensor & out) {
+    return at::_ops::_cudnn_rnn_flatten_weight_out::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & _cudnn_rnn_flatten_weight_outf(at::TensorList weight_arr, int64_t weight_stride0, c10::SymInt input_size, int64_t mode, c10::SymInt hidden_size, c10::SymInt proj_size, int64_t num_layers, bool batch_first, bool bidirectional, at::Tensor & out) {
+    return at::_ops::_cudnn_rnn_flatten_weight_out::call(weight_arr, weight_stride0, input_size, mode, hidden_size, proj_size, num_layers, batch_first, bidirectional, out);
+  }
+}
+
+}

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

@@ -0,0 +1,22 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor & _cudnn_rnn_flatten_weight_out_symint(at::TensorList weight_arr, int64_t weight_stride0, c10::SymInt input_size, int64_t mode, c10::SymInt hidden_size, c10::SymInt proj_size, int64_t num_layers, bool batch_first, bool bidirectional, at::Tensor & out);
+TORCH_API at::Tensor _cudnn_rnn_flatten_weight(at::TensorList weight_arr, int64_t weight_stride0, int64_t input_size, int64_t mode, int64_t hidden_size, int64_t proj_size, int64_t num_layers, bool batch_first, bool bidirectional);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,30 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/_cufft_set_plan_cache_max_size_ops.h>
+
+namespace at {
+
+
+// aten::_cufft_set_plan_cache_max_size(DeviceIndex device_index, int max_size) -> ()
+inline void _cufft_set_plan_cache_max_size(at::DeviceIndex device_index, int64_t max_size) {
+    return at::_ops::_cufft_set_plan_cache_max_size::call(device_index, max_size);
+}
+
+}

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

@@ -0,0 +1,21 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor _fw_primal(const at::Tensor & self, int64_t level);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _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_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_masked_softmax")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(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_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_masked_softmax")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(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

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

@@ -0,0 +1,31 @@
+#pragma once
+
+// @generated by torchgen/gen.py 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> _native_batch_norm_legit_functional(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, bool training, double momentum, double eps);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _batch_norm_legit_cpu_out(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, at::Tensor & running_mean, 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_legit_cpu(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, at::Tensor & running_mean, at::Tensor & running_var, bool training, double momentum, double eps);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _batch_norm_legit_cuda_out(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, at::Tensor & running_mean, 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_legit_cuda(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, at::Tensor & running_mean, at::Tensor & running_var, bool training, double momentum, double eps);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _mkldnn_batch_norm_legit(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, at::Tensor & running_mean, at::Tensor & running_var, bool training, double momentum, double eps);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _batch_norm_legit_no_stats_cpu(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, bool training, double momentum, double eps);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _batch_norm_legit_no_stats_cpu_out(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, 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_legit_no_stats_cuda(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, bool training, double momentum, double eps);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _batch_norm_legit_no_stats_cuda_out(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, 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_legit_no_stats(const at::Tensor & input, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & bias, bool training, double momentum, double eps);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,20 @@
+#pragma once
+
+// @generated by torchgen/gen.py 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

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

@@ -0,0 +1,28 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _nested_select_backward {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, int64_t, c10::SymInt);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_nested_select_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_nested_select_backward(Tensor grad_output, Tensor self, int dim, SymInt index) -> Tensor")
+  static at::Tensor call(const at::Tensor & grad_output, const at::Tensor & self, int64_t dim, c10::SymInt index);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & self, int64_t dim, c10::SymInt index);
+};
+
+}} // namespace at::_ops

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

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/_new_zeros_with_same_feature_meta_ops.h>
+
+namespace at {
+
+
+// aten::_new_zeros_with_same_feature_meta(Tensor self, Tensor other, *, int self_num_batch_dims=0) -> Tensor
+inline at::Tensor _new_zeros_with_same_feature_meta(const at::Tensor & self, const at::Tensor & other, int64_t self_num_batch_dims=0) {
+    return at::_ops::_new_zeros_with_same_feature_meta::call(self, other, self_num_batch_dims);
+}
+
+// aten::_new_zeros_with_same_feature_meta.out(Tensor self, Tensor other, *, int self_num_batch_dims=0, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _new_zeros_with_same_feature_meta_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other, int64_t self_num_batch_dims=0) {
+    return at::_ops::_new_zeros_with_same_feature_meta_out::call(self, other, self_num_batch_dims, out);
+}
+// aten::_new_zeros_with_same_feature_meta.out(Tensor self, Tensor other, *, int self_num_batch_dims=0, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _new_zeros_with_same_feature_meta_outf(const at::Tensor & self, const at::Tensor & other, int64_t self_num_batch_dims, at::Tensor & out) {
+    return at::_ops::_new_zeros_with_same_feature_meta_out::call(self, other, self_num_batch_dims, out);
+}
+
+}

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

@@ -0,0 +1,28 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _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_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_thnn_differentiable_gru_cell_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_bilinear2d_aa_meta.h

@@ -0,0 +1,27 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeMetaFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/TensorIterator.h>
+#include <ATen/TensorMeta.h>
+#include <tuple>
+#include <vector>
+
+namespace at {
+namespace meta {
+
+struct TORCH_API structured__upsample_bilinear2d_aa : public at::impl::MetaBase {
+    
+    
+    void meta(const at::Tensor & self, at::ArrayRef<int64_t> output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w);
+};
+
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_bilinear2d_aa_native.h

@@ -0,0 +1,27 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+#include <ATen/ops/_upsample_bilinear2d_aa_meta.h>
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor _upsample_bilinear2d_aa(const at::Tensor & input, at::OptionalIntArrayRef output_size, bool align_corners, ::std::optional<at::ArrayRef<double>> scale_factors);
+struct TORCH_API structured__upsample_bilinear2d_aa_out_cpu : public at::meta::structured__upsample_bilinear2d_aa {
+void impl(const at::Tensor & self, at::ArrayRef<int64_t> output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, const at::Tensor & out);
+};
+struct TORCH_API structured__upsample_bilinear2d_aa_out_cuda : public at::meta::structured__upsample_bilinear2d_aa {
+void impl(const at::Tensor & self, at::ArrayRef<int64_t> output_size, bool align_corners, ::std::optional<double> scales_h, ::std::optional<double> scales_w, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,35 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/_use_cudnn_ctc_loss_ops.h>
+
+namespace at {
+
+
+// aten::_use_cudnn_ctc_loss(Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, int blank) -> bool
+inline bool _use_cudnn_ctc_loss(const at::Tensor & log_probs, const at::Tensor & targets, at::IntArrayRef input_lengths, at::IntArrayRef target_lengths, int64_t blank) {
+    return at::_ops::_use_cudnn_ctc_loss::call(log_probs, targets, input_lengths, target_lengths, blank);
+}
+
+// aten::_use_cudnn_ctc_loss.Tensor(Tensor log_probs, Tensor targets, Tensor input_lengths, Tensor target_lengths, int blank) -> bool
+inline bool _use_cudnn_ctc_loss(const at::Tensor & log_probs, const at::Tensor & targets, const at::Tensor & input_lengths, const at::Tensor & target_lengths, int64_t blank) {
+    return at::_ops::_use_cudnn_ctc_loss_Tensor::call(log_probs, targets, input_lengths, target_lengths, blank);
+}
+
+}

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

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

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

@@ -0,0 +1,21 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <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);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/adaptive_avg_pool3d_compositeimplicitautograd_dispatch.h

@@ -0,0 +1,24 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor adaptive_avg_pool3d(const at::Tensor & self, at::IntArrayRef output_size);
+TORCH_API at::Tensor adaptive_avg_pool3d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size);
+
+} // namespace compositeimplicitautograd
+} // namespace at

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

@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API at::Tensor 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

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

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/bitwise_not_ops.h>
+
+namespace at {
+
+
+// aten::bitwise_not(Tensor self) -> Tensor
+inline at::Tensor bitwise_not(const at::Tensor & self) {
+    return at::_ops::bitwise_not::call(self);
+}
+
+// aten::bitwise_not.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & bitwise_not_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::bitwise_not_out::call(self, out);
+}
+// aten::bitwise_not.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & bitwise_not_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::bitwise_not_out::call(self, out);
+}
+
+}

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

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API bmm {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::bmm")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "bmm(Tensor self, Tensor mat2) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & mat2);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat2);
+};
+
+struct TORCH_API bmm_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::bmm")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "bmm.out(Tensor self, Tensor mat2, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out);
+};
+
+}} // namespace at::_ops

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

@@ -0,0 +1,24 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor & complex_out(at::Tensor & out, const at::Tensor & real, const at::Tensor & imag);
+TORCH_API at::Tensor & complex_outf(const at::Tensor & real, const at::Tensor & imag, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at

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

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API complex {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::complex")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "complex(Tensor real, Tensor imag) -> Tensor")
+  static at::Tensor call(const at::Tensor & real, const at::Tensor & imag);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & real, const at::Tensor & imag);
+};
+
+struct TORCH_API complex_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::complex")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "complex.out(Tensor real, Tensor imag, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & real, const at::Tensor & imag, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & real, const at::Tensor & imag, at::Tensor & out);
+};
+
+}} // namespace at::_ops

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

@@ -0,0 +1,28 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API ::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);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> convolution_backward_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, 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_outf(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, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> convolution_backward_symint_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, 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);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> convolution_backward_symint_outf(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 compositeexplicitautograd
+} // namespace at

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

@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API at::Tensor 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 cpu
+} // namespace at

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

@@ -0,0 +1,22 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <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_affine_grid_generator_backward_out(const at::Tensor & grad, int64_t N, int64_t C, int64_t H, int64_t W, at::Tensor & out);
+TORCH_API at::Tensor cudnn_affine_grid_generator_backward(const at::Tensor & grad, int64_t N, int64_t C, int64_t H, int64_t W);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API cudnn_batch_norm {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, bool, double, double);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::cudnn_batch_norm")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "cudnn_batch_norm(Tensor input, Tensor weight, Tensor? bias, Tensor? running_mean, Tensor? running_var, bool training, float exponential_average_factor, float epsilon) -> (Tensor, Tensor, Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> call(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);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, 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);
+};
+
+struct TORCH_API cudnn_batch_norm_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, bool, double, double, at::Tensor &, at::Tensor &, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::cudnn_batch_norm")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "cudnn_batch_norm.out(Tensor input, Tensor weight, Tensor? bias, Tensor? running_mean, Tensor? running_var, bool training, float exponential_average_factor, float epsilon, *, 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 & 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);
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, 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 at::_ops

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

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

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

@@ -0,0 +1,27 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeMetaFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/TensorIterator.h>
+#include <ATen/TensorMeta.h>
+#include <tuple>
+#include <vector>
+
+namespace at {
+namespace meta {
+
+struct TORCH_API structured_exp : public TensorIteratorBase {
+    
+    
+    void meta(const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,28 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API fake_quantize_per_channel_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_STR_INL_EXCEPT_WIN_CUDA(name, "aten::fake_quantize_per_channel_affine_cachemask_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "fake_quantize_per_channel_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

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

@@ -0,0 +1,28 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API fbgemm_linear_quantize_weight {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,double,int64_t> (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::fbgemm_linear_quantize_weight")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "fbgemm_linear_quantize_weight(Tensor input) -> (Tensor, Tensor, float, int)")
+  static ::std::tuple<at::Tensor,at::Tensor,double,int64_t> call(const at::Tensor & input);
+  static ::std::tuple<at::Tensor,at::Tensor,double,int64_t> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input);
+};
+
+}} // namespace at::_ops

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

@@ -0,0 +1,23 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor fix(const at::Tensor & self);
+TORCH_API at::Tensor & fix_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & fix_(at::Tensor & self);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fractional_max_pool3d_ops.h

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API fractional_max_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_STR_INL_EXCEPT_WIN_CUDA(name, "aten::fractional_max_pool3d")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "output")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(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_STR_INL_EXCEPT_WIN_CUDA(name, "aten::fractional_max_pool3d")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(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

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

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

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

@@ -0,0 +1,21 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor instance_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 use_input_stats, double momentum, double eps, bool cudnn_enabled);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/isnan_ops.h>
+
+namespace at {
+
+
+// aten::isnan(Tensor self) -> Tensor
+inline at::Tensor isnan(const at::Tensor & self) {
+    return at::_ops::isnan::call(self);
+}
+
+// aten::isnan.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & isnan_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::isnan_out::call(self, out);
+}
+// aten::isnan.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & isnan_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::isnan_out::call(self, out);
+}
+
+}

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

@@ -0,0 +1,47 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/layer_norm_ops.h>
+
+namespace at {
+
+
+// aten::layer_norm(Tensor input, SymInt[] normalized_shape, Tensor? weight=None, Tensor? bias=None, float eps=1e-05, bool cudnn_enable=True) -> Tensor
+inline at::Tensor layer_norm(const at::Tensor & input, at::IntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight={}, const ::std::optional<at::Tensor> & bias={}, double eps=1e-05, bool cudnn_enable=true) {
+    return at::_ops::layer_norm::call(input, c10::fromIntArrayRefSlow(normalized_shape), weight, bias, eps, cudnn_enable);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor layer_norm(const at::Tensor & input, at::IntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight={}, const ::std::optional<at::Tensor> & bias={}, double eps=1e-05, bool cudnn_enable=true) {
+    return at::_ops::layer_norm::call(input, c10::fromIntArrayRefSlow(normalized_shape), weight, bias, eps, cudnn_enable);
+  }
+}
+
+// aten::layer_norm(Tensor input, SymInt[] normalized_shape, Tensor? weight=None, Tensor? bias=None, float eps=1e-05, bool cudnn_enable=True) -> Tensor
+inline at::Tensor 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=1e-05, bool cudnn_enable=true) {
+    return at::_ops::layer_norm::call(input, normalized_shape, weight, bias, eps, cudnn_enable);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor layer_norm(const at::Tensor & input, c10::SymIntArrayRef normalized_shape, const ::std::optional<at::Tensor> & weight={}, const ::std::optional<at::Tensor> & bias={}, double eps=1e-05, bool cudnn_enable=true) {
+    return at::_ops::layer_norm::call(input, normalized_shape, weight, bias, eps, cudnn_enable);
+  }
+}
+
+}

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

@@ -0,0 +1,26 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+#include <ATen/ops/leaky_relu_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_leaky_relu_out : public at::meta::structured_leaky_relu {
+void impl(const at::Tensor & self, const at::Scalar & negative_slope, const at::Tensor & out);
+};
+TORCH_API at::Tensor leaky_relu_quantized_cpu(const at::Tensor & self, const at::Scalar & negative_slope=0.01);
+TORCH_API at::Tensor & leaky_relu_out_quantized_cpu(const at::Tensor & self, const at::Scalar & negative_slope, at::Tensor & out);
+TORCH_API at::Tensor & leaky_relu_quantized_cpu_(at::Tensor & self, const at::Scalar & negative_slope=0.01);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,22 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor linalg_matrix_power(const at::Tensor & self, int64_t n);
+TORCH_API at::Tensor & linalg_matrix_power_out(const at::Tensor & self, int64_t n, at::Tensor & out);
+} // namespace native
+} // namespace at

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

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

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

@@ -0,0 +1,24 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API at::Tensor & linspace_out(at::Tensor & out, const at::Scalar & start, const at::Scalar & end, int64_t steps);
+TORCH_API at::Tensor & linspace_outf(const at::Scalar & start, const at::Scalar & end, int64_t steps, at::Tensor & out);
+
+} // namespace meta
+} // namespace at

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

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

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

@@ -0,0 +1,35 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/meshgrid_ops.h>
+
+namespace at {
+
+
+// aten::meshgrid(Tensor[] tensors) -> Tensor[]
+inline ::std::vector<at::Tensor> meshgrid(at::TensorList tensors) {
+    return at::_ops::meshgrid::call(tensors);
+}
+
+// aten::meshgrid.indexing(Tensor[] tensors, *, str indexing) -> Tensor[]
+inline ::std::vector<at::Tensor> meshgrid(at::TensorList tensors, c10::string_view indexing) {
+    return at::_ops::meshgrid_indexing::call(tensors, indexing);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mkldnn_max_pool2d.h

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/mkldnn_max_pool2d_ops.h>
+
+namespace at {
+
+
+// aten::mkldnn_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 mkldnn_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::mkldnn_max_pool2d::call(self, kernel_size, stride, padding, dilation, ceil_mode);
+}
+
+// aten::mkldnn_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 & mkldnn_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::mkldnn_max_pool2d_out::call(self, kernel_size, stride, padding, dilation, ceil_mode, out);
+}
+// aten::mkldnn_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 & mkldnn_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::mkldnn_max_pool2d_out::call(self, kernel_size, stride, padding, dilation, ceil_mode, out);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mkldnn_reorder_conv3d_weight_compositeexplicitautograd_dispatch.h

@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & mkldnn_reorder_conv3d_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);
+TORCH_API at::Tensor & mkldnn_reorder_conv3d_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);
+TORCH_API at::Tensor & mkldnn_reorder_conv3d_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);
+TORCH_API at::Tensor & mkldnn_reorder_conv3d_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);
+
+} // namespace compositeexplicitautograd
+} // namespace at

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

@@ -0,0 +1,22 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor movedim(const at::Tensor & self, at::IntArrayRef source, at::IntArrayRef destination);
+TORCH_API at::Tensor movedim(const at::Tensor & self, int64_t source, int64_t destination);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,28 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API ::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 cuda
+} // namespace at

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

@@ -0,0 +1,21 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor pad_symint(const at::Tensor & self, c10::SymIntArrayRef pad, c10::string_view mode="constant", ::std::optional<double> value=::std::nullopt);
+} // namespace native
+} // namespace at

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/relu6_compositeimplicitautograd_dispatch.h

@@ -0,0 +1,24 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor relu6(const at::Tensor & self);
+TORCH_API at::Tensor & relu6_(at::Tensor & self);
+
+} // namespace compositeimplicitautograd
+} // namespace at

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

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API select_copy_int {
+  using schema = at::Tensor (const at::Tensor &, int64_t, c10::SymInt);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::select_copy")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "int")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "select_copy.int(Tensor self, int dim, SymInt index) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, int64_t dim, c10::SymInt index);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, c10::SymInt index);
+};
+
+struct TORCH_API select_copy_int_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, c10::SymInt, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::select_copy")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "int_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "select_copy.int_out(Tensor self, int dim, SymInt index, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, int64_t dim, c10::SymInt index, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, c10::SymInt index, at::Tensor & out);
+};
+
+}} // namespace at::_ops

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

@@ -0,0 +1,27 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeMetaFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/TensorIterator.h>
+#include <ATen/TensorMeta.h>
+#include <tuple>
+#include <vector>
+
+namespace at {
+namespace meta {
+
+struct TORCH_API structured_sigmoid_backward : public TensorIteratorBase {
+    
+    
+    void meta(const at::Tensor & grad_output, const at::Tensor & output);
+};
+
+} // namespace native
+} // namespace at