Add files using upload-large-folder tool

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

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

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_copy_from_and_resize.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/_copy_from_and_resize_ops.h>
+
+namespace at {
+
+
+// aten::_copy_from_and_resize(Tensor self, Tensor dst) -> Tensor
+inline at::Tensor _copy_from_and_resize(const at::Tensor & self, const at::Tensor & dst) {
+    return at::_ops::_copy_from_and_resize::call(self, dst);
+}
+
+// aten::_copy_from_and_resize.out(Tensor self, Tensor dst, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _copy_from_and_resize_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & dst) {
+    return at::_ops::_copy_from_and_resize_out::call(self, dst, out);
+}
+// aten::_copy_from_and_resize.out(Tensor self, Tensor dst, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _copy_from_and_resize_outf(const at::Tensor & self, const at::Tensor & dst, at::Tensor & out) {
+    return at::_ops::_copy_from_and_resize_out::call(self, dst, out);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_debug_has_internal_overlap_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 int64_t _debug_has_internal_overlap(const at::Tensor & self);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/_embedding_bag_per_sample_weights_backward_ops.h>
+
+namespace at {
+
+
+// aten::_embedding_bag_per_sample_weights_backward(Tensor grad, Tensor weight, Tensor indices, Tensor offsets, Tensor offset2bag, int mode, int padding_idx=-1) -> Tensor
+inline at::Tensor _embedding_bag_per_sample_weights_backward(const at::Tensor & grad, const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, const at::Tensor & offset2bag, int64_t mode, int64_t padding_idx=-1) {
+    return at::_ops::_embedding_bag_per_sample_weights_backward::call(grad, weight, indices, offsets, offset2bag, mode, padding_idx);
+}
+
+// aten::_embedding_bag_per_sample_weights_backward.out(Tensor grad, Tensor weight, Tensor indices, Tensor offsets, Tensor offset2bag, int mode, int padding_idx=-1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _embedding_bag_per_sample_weights_backward_out(at::Tensor & out, const at::Tensor & grad, const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, const at::Tensor & offset2bag, int64_t mode, int64_t padding_idx=-1) {
+    return at::_ops::_embedding_bag_per_sample_weights_backward_out::call(grad, weight, indices, offsets, offset2bag, mode, padding_idx, out);
+}
+// aten::_embedding_bag_per_sample_weights_backward.out(Tensor grad, Tensor weight, Tensor indices, Tensor offsets, Tensor offset2bag, int mode, int padding_idx=-1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _embedding_bag_per_sample_weights_backward_outf(const at::Tensor & grad, const at::Tensor & weight, const at::Tensor & indices, const at::Tensor & offsets, const at::Tensor & offset2bag, int64_t mode, int64_t padding_idx, at::Tensor & out) {
+    return at::_ops::_embedding_bag_per_sample_weights_backward_out::call(grad, weight, indices, offsets, offset2bag, mode, padding_idx, out);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_flash_attention_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 _flash_attention_backward {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,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 &, c10::SymInt, c10::SymInt, double, bool, const at::Tensor &, const at::Tensor &, ::std::optional<double>, ::std::optional<c10::SymInt>, ::std::optional<c10::SymInt>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_flash_attention_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_flash_attention_backward(Tensor grad_out, Tensor query, Tensor key, Tensor value, Tensor out, Tensor logsumexp, Tensor cum_seq_q, Tensor cum_seq_k, SymInt max_q, SymInt max_k, float dropout_p, bool is_causal, Tensor philox_seed, Tensor philox_offset, *, float? scale=None, SymInt? window_size_left=None, SymInt? window_size_right=None) -> (Tensor, Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> call(const at::Tensor & grad_out, const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const at::Tensor & out, const at::Tensor & logsumexp, const at::Tensor & cum_seq_q, const at::Tensor & cum_seq_k, c10::SymInt max_q, c10::SymInt max_k, double dropout_p, bool is_causal, const at::Tensor & philox_seed, const at::Tensor & philox_offset, ::std::optional<double> scale, ::std::optional<c10::SymInt> window_size_left, ::std::optional<c10::SymInt> window_size_right);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_out, const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const at::Tensor & out, const at::Tensor & logsumexp, const at::Tensor & cum_seq_q, const at::Tensor & cum_seq_k, c10::SymInt max_q, c10::SymInt max_k, double dropout_p, bool is_causal, const at::Tensor & philox_seed, const at::Tensor & philox_offset, ::std::optional<double> scale, ::std::optional<c10::SymInt> window_size_left, ::std::optional<c10::SymInt> window_size_right);
+};
+
+}} // namespace at::_ops

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_functional_assert_async_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 _functional_assert_async(const at::Tensor & self, c10::string_view assert_msg, const at::Tensor & dep_token);
+
+} // namespace cpu
+} // namespace at

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

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_linalg_solve_ex_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__linalg_solve_ex : public at::impl::MetaBase {
+    
+    
+    void meta(const at::Tensor & A, const at::Tensor & B, bool left, bool check_errors);
+};
+
+} // namespace native
+} // namespace at

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

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

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_reshape_alias.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/_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<T, int64_t>::value>>
+  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<T, c10::SymInt>::value>>
+  at::Tensor _reshape_alias(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride) {
+    return at::_ops::_reshape_alias::call(self, size, stride);
+  }
+}
+
+}

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_sparse_semi_structured_linear_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 _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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d.h

@@ -0,0 +1,113 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/_upsample_nearest_exact1d_ops.h>
+
+namespace at {
+
+
+// aten::_upsample_nearest_exact1d.vec(Tensor input, SymInt[]? output_size, float[]? scale_factors) -> Tensor
+inline at::Tensor _upsample_nearest_exact1d(const at::Tensor & input, at::OptionalIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::_upsample_nearest_exact1d_vec::call(input, output_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*output_size)) : ::std::nullopt, scale_factors);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor _upsample_nearest_exact1d(const at::Tensor & input, at::OptionalIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::_upsample_nearest_exact1d_vec::call(input, output_size.has_value() ? ::std::make_optional(c10::fromIntArrayRefSlow(*output_size)) : ::std::nullopt, scale_factors);
+  }
+}
+
+// aten::_upsample_nearest_exact1d.vec(Tensor input, SymInt[]? output_size, float[]? scale_factors) -> Tensor
+inline at::Tensor _upsample_nearest_exact1d_symint(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::_upsample_nearest_exact1d_vec::call(input, output_size, scale_factors);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor _upsample_nearest_exact1d(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, ::std::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::_upsample_nearest_exact1d_vec::call(input, output_size, scale_factors);
+  }
+}
+
+// aten::_upsample_nearest_exact1d.out(Tensor self, SymInt[1] output_size, float? scales=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _upsample_nearest_exact1d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales=::std::nullopt) {
+    return at::_ops::_upsample_nearest_exact1d_out::call(self, c10::fromIntArrayRefSlow(output_size), scales, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & _upsample_nearest_exact1d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales=::std::nullopt) {
+    return at::_ops::_upsample_nearest_exact1d_out::call(self, c10::fromIntArrayRefSlow(output_size), scales, out);
+  }
+}
+
+// aten::_upsample_nearest_exact1d.out(Tensor self, SymInt[1] output_size, float? scales=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _upsample_nearest_exact1d_outf(const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales, at::Tensor & out) {
+    return at::_ops::_upsample_nearest_exact1d_out::call(self, c10::fromIntArrayRefSlow(output_size), scales, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & _upsample_nearest_exact1d_outf(const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales, at::Tensor & out) {
+    return at::_ops::_upsample_nearest_exact1d_out::call(self, c10::fromIntArrayRefSlow(output_size), scales, out);
+  }
+}
+
+// aten::_upsample_nearest_exact1d.out(Tensor self, SymInt[1] output_size, float? scales=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _upsample_nearest_exact1d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales=::std::nullopt) {
+    return at::_ops::_upsample_nearest_exact1d_out::call(self, output_size, scales, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & _upsample_nearest_exact1d_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales=::std::nullopt) {
+    return at::_ops::_upsample_nearest_exact1d_out::call(self, output_size, scales, out);
+  }
+}
+
+// aten::_upsample_nearest_exact1d.out(Tensor self, SymInt[1] output_size, float? scales=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _upsample_nearest_exact1d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales, at::Tensor & out) {
+    return at::_ops::_upsample_nearest_exact1d_out::call(self, output_size, scales, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & _upsample_nearest_exact1d_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales, at::Tensor & out) {
+    return at::_ops::_upsample_nearest_exact1d_out::call(self, output_size, scales, out);
+  }
+}
+
+// aten::_upsample_nearest_exact1d(Tensor self, SymInt[1] output_size, float? scales=None) -> Tensor
+inline at::Tensor _upsample_nearest_exact1d(const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales=::std::nullopt) {
+    return at::_ops::_upsample_nearest_exact1d::call(self, c10::fromIntArrayRefSlow(output_size), scales);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor _upsample_nearest_exact1d(const at::Tensor & self, at::IntArrayRef output_size, ::std::optional<double> scales=::std::nullopt) {
+    return at::_ops::_upsample_nearest_exact1d::call(self, c10::fromIntArrayRefSlow(output_size), scales);
+  }
+}
+
+// aten::_upsample_nearest_exact1d(Tensor self, SymInt[1] output_size, float? scales=None) -> Tensor
+inline at::Tensor _upsample_nearest_exact1d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales=::std::nullopt) {
+    return at::_ops::_upsample_nearest_exact1d::call(self, output_size, scales);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor _upsample_nearest_exact1d(const at::Tensor & self, c10::SymIntArrayRef output_size, ::std::optional<double> scales=::std::nullopt) {
+    return at::_ops::_upsample_nearest_exact1d::call(self, output_size, scales);
+  }
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_backward_compositeexplicitautogradnonfunctional_dispatch.h

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_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_nearest_exact3d : public at::impl::MetaBase {
+    
+    
+    void meta(const at::Tensor & self, at::ArrayRef<int64_t> output_size, ::std::optional<double> scales_d, ::std::optional<double> scales_h, ::std::optional<double> scales_w);
+};
+
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/adaptive_max_pool2d_backward_cpu_dispatch.h

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

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

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

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/batch_norm_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 batch_norm_backward {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, const ::std::optional<at::Tensor> &, bool, double, ::std::array<bool,3>, 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::batch_norm_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "batch_norm_backward(Tensor grad_out, Tensor input, Tensor weight, Tensor? running_mean, Tensor? running_var, Tensor? save_mean, Tensor? save_var, bool update, float eps, bool[3] output_mask, Tensor reserve) -> (Tensor, Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> call(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);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, 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 at::_ops

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/ccol_indices_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 ccol_indices {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::ccol_indices")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/convolution_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 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_STR_INL_EXCEPT_WIN_CUDA(name, "aten::convolution")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(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_STR_INL_EXCEPT_WIN_CUDA(name, "aten::convolution")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(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

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

@@ -0,0 +1,25 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor & count_nonzero_dim_IntList_out(const at::Tensor & self, at::IntArrayRef dim, at::Tensor & out);
+TORCH_API at::Tensor count_nonzero_cpu(const at::Tensor & self, at::IntArrayRef dim);
+TORCH_API at::Tensor count_nonzero_cuda(const at::Tensor & self, at::IntArrayRef dim);
+TORCH_API at::Tensor count_nonzero(const at::Tensor & self, ::std::optional<int64_t> dim=::std::nullopt);
+TORCH_API at::Tensor & count_nonzero_out(const at::Tensor & self, ::std::optional<int64_t> dim, at::Tensor & out);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/crow_indices_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 crow_indices {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::crow_indices")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "crow_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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/cudnn_convolution_relu_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_convolution_relu_out_symint(const at::Tensor & self, const at::Tensor & weight, 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_relu(const at::Tensor & self, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, int64_t groups);
+} // namespace native
+} // namespace at

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

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

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

@@ -0,0 +1,53 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#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);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/diag_embed_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 & diag_embed_out(const at::Tensor & self, int64_t offset, int64_t dim1, int64_t dim2, at::Tensor & out);
+TORCH_API at::Tensor diag_embed(const at::Tensor & self, int64_t offset=0, int64_t dim1=-2, int64_t dim2=-1);
+} // namespace native
+} // namespace at

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

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

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

@@ -0,0 +1,63 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/fill_ops.h>
+
+namespace at {
+
+
+// aten::fill.Scalar(Tensor self, Scalar value) -> Tensor
+inline at::Tensor fill(const at::Tensor & self, const at::Scalar & value) {
+    return at::_ops::fill_Scalar::call(self, value);
+}
+
+// aten::fill.Tensor(Tensor self, Tensor value) -> Tensor
+inline at::Tensor fill(const at::Tensor & self, const at::Tensor & value) {
+    return at::_ops::fill_Tensor::call(self, value);
+}
+
+// aten::fill_.Scalar(Tensor(a!) self, Scalar value) -> Tensor(a!)
+inline at::Tensor & fill_(at::Tensor & self, const at::Scalar & value) {
+    return at::_ops::fill__Scalar::call(self, value);
+}
+
+// aten::fill_.Tensor(Tensor(a!) self, Tensor value) -> Tensor(a!)
+inline at::Tensor & fill_(at::Tensor & self, const at::Tensor & value) {
+    return at::_ops::fill__Tensor::call(self, value);
+}
+
+// aten::fill.Scalar_out(Tensor self, Scalar value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fill_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & value) {
+    return at::_ops::fill_Scalar_out::call(self, value, out);
+}
+// aten::fill.Scalar_out(Tensor self, Scalar value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fill_outf(const at::Tensor & self, const at::Scalar & value, at::Tensor & out) {
+    return at::_ops::fill_Scalar_out::call(self, value, out);
+}
+
+// aten::fill.Tensor_out(Tensor self, Tensor value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fill_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & value) {
+    return at::_ops::fill_Tensor_out::call(self, value, out);
+}
+// aten::fill.Tensor_out(Tensor self, Tensor value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fill_outf(const at::Tensor & self, const at::Tensor & value, at::Tensor & out) {
+    return at::_ops::fill_Tensor_out::call(self, value, out);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fractional_max_pool2d_cpu_dispatch.h

@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> fractional_max_pool2d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples);
+TORCH_API ::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);
+TORCH_API ::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);
+
+} // namespace cpu
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fractional_max_pool3d_backward_cuda_dispatch.h

@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor fractional_max_pool3d_backward(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(at::Tensor & grad_input, 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_outf(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 cuda
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/glu_backward_jvp_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 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 cpu
+} // namespace at

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

@@ -0,0 +1,44 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/hardtanh_ops.h>
+
+namespace at {
+
+
+// aten::hardtanh.out(Tensor self, Scalar min_val=-1, Scalar max_val=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & hardtanh_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & min_val=-1, const at::Scalar & max_val=1) {
+    return at::_ops::hardtanh_out::call(self, min_val, max_val, out);
+}
+// aten::hardtanh.out(Tensor self, Scalar min_val=-1, Scalar max_val=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & hardtanh_outf(const at::Tensor & self, const at::Scalar & min_val, const at::Scalar & max_val, at::Tensor & out) {
+    return at::_ops::hardtanh_out::call(self, min_val, max_val, out);
+}
+
+// aten::hardtanh(Tensor self, Scalar min_val=-1, Scalar max_val=1) -> Tensor
+inline at::Tensor hardtanh(const at::Tensor & self, const at::Scalar & min_val=-1, const at::Scalar & max_val=1) {
+    return at::_ops::hardtanh::call(self, min_val, max_val);
+}
+
+// aten::hardtanh_(Tensor(a!) self, Scalar min_val=-1, Scalar max_val=1) -> Tensor(a!)
+inline at::Tensor & hardtanh_(at::Tensor & self, const at::Scalar & min_val=-1, const at::Scalar & max_val=1) {
+    return at::_ops::hardtanh_::call(self, min_val, max_val);
+}
+
+}

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

@@ -0,0 +1,53 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/index_select_ops.h>
+
+namespace at {
+
+
+// aten::index_select.out(Tensor self, int dim, Tensor index, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & index_select_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index) {
+    return at::_ops::index_select_out::call(self, dim, index, out);
+}
+// aten::index_select.out(Tensor self, int dim, Tensor index, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & index_select_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, at::Tensor & out) {
+    return at::_ops::index_select_out::call(self, dim, index, out);
+}
+
+// aten::index_select(Tensor self, int dim, Tensor index) -> Tensor
+inline at::Tensor index_select(const at::Tensor & self, int64_t dim, const at::Tensor & index) {
+    return at::_ops::index_select::call(self, dim, index);
+}
+
+// aten::index_select.dimname_out(Tensor self, Dimname dim, Tensor index, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & index_select_out(at::Tensor & out, const at::Tensor & self, at::Dimname dim, const at::Tensor & index) {
+    return at::_ops::index_select_dimname_out::call(self, dim, index, out);
+}
+// aten::index_select.dimname_out(Tensor self, Dimname dim, Tensor index, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & index_select_outf(const at::Tensor & self, at::Dimname dim, const at::Tensor & index, at::Tensor & out) {
+    return at::_ops::index_select_dimname_out::call(self, dim, index, out);
+}
+
+// aten::index_select.dimname(Tensor self, Dimname dim, Tensor index) -> Tensor
+inline at::Tensor index_select(const at::Tensor & self, at::Dimname dim, const at::Tensor & index) {
+    return at::_ops::index_select_dimname::call(self, dim, index);
+}
+
+}

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/isposinf_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_isposinf : public TensorIteratorBase {
+    
+    
+    void meta(const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at

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

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

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_ldl_factor_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 ::std::tuple<at::Tensor,at::Tensor> linalg_ldl_factor(const at::Tensor & self, bool hermitian=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> linalg_ldl_factor_out(at::Tensor & LD, at::Tensor & pivots, const at::Tensor & self, bool hermitian=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> linalg_ldl_factor_outf(const at::Tensor & self, bool hermitian, at::Tensor & LD, at::Tensor & pivots);
+
+} // namespace compositeimplicitautograd
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_vecdot.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/linalg_vecdot_ops.h>
+
+namespace at {
+
+
+// aten::linalg_vecdot(Tensor x, Tensor y, *, int dim=-1) -> Tensor
+inline at::Tensor linalg_vecdot(const at::Tensor & x, const at::Tensor & y, int64_t dim=-1) {
+    return at::_ops::linalg_vecdot::call(x, y, dim);
+}
+
+// aten::linalg_vecdot.out(Tensor x, Tensor y, *, int dim=-1, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_vecdot_out(at::Tensor & out, const at::Tensor & x, const at::Tensor & y, int64_t dim=-1) {
+    return at::_ops::linalg_vecdot_out::call(x, y, dim, out);
+}
+// aten::linalg_vecdot.out(Tensor x, Tensor y, *, int dim=-1, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_vecdot_outf(const at::Tensor & x, const at::Tensor & y, int64_t dim, at::Tensor & out) {
+    return at::_ops::linalg_vecdot_out::call(x, y, dim, out);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linear_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 linear(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias={});
+TORCH_API at::Tensor & linear_out(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias, at::Tensor & out);
+TORCH_API at::Tensor nested_linear(const at::Tensor & input, const at::Tensor & weight, const ::std::optional<at::Tensor> & bias={});
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/log_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_log : public TensorIteratorBase {
+    
+    
+    void meta(const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/miopen_depthwise_convolution_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 miopen_depthwise_convolution {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymInt, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::miopen_depthwise_convolution")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "miopen_depthwise_convolution(Tensor self, Tensor weight, Tensor? bias, SymInt[] padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic) -> Tensor")
+  static at::Tensor call(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);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, 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);
+};
+
+struct TORCH_API miopen_depthwise_convolution_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, const ::std::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymInt, bool, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::miopen_depthwise_convolution")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "miopen_depthwise_convolution.out(Tensor self, Tensor weight, Tensor? bias, SymInt[] padding, SymInt[] stride, SymInt[] dilation, SymInt groups, bool benchmark, bool deterministic, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(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, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, 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, at::Tensor & out);
+};
+
+}} // namespace at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mm_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 mm(const at::Tensor & self, const at::Tensor & mat2);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/nanmean_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 nanmean {
+  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_STR_INL_EXCEPT_WIN_CUDA(name, "aten::nanmean")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "nanmean(Tensor self, int[1]? dim=None, 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 nanmean_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_STR_INL_EXCEPT_WIN_CUDA(name, "aten::nanmean")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "nanmean.out(Tensor self, int[1]? dim=None, 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);
+};
+
+}} // namespace at::_ops