Add files using upload-large-folder tool

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_add_batch_dim_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 _add_batch_dim(const at::Tensor & self, int64_t batch_dim, int64_t level);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_amp_update_scale_meta_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 meta {
+
+TORCH_API at::Tensor & _amp_update_scale_(at::Tensor & self, at::Tensor & growth_tracker, const at::Tensor & found_inf, double scale_growth_factor, double scale_backoff_factor, int64_t growth_interval);
+
+} // namespace meta
+} // namespace at

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

@@ -0,0 +1,30 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/_batch_norm_impl_index_backward_ops.h>
+
+namespace at {
+
+
+// aten::_batch_norm_impl_index_backward(int impl_index, Tensor input, Tensor grad_output, Tensor? weight, Tensor? running_mean, Tensor? running_var, Tensor? save_mean, Tensor? save_var_transform, bool train, float eps, bool[3] output_mask, Tensor reservedSpace) -> (Tensor, Tensor, Tensor)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _batch_norm_impl_index_backward(int64_t impl_index, const at::Tensor & input, const at::Tensor & grad_output, const ::std::optional<at::Tensor> & weight, const ::std::optional<at::Tensor> & running_mean, const ::std::optional<at::Tensor> & running_var, const ::std::optional<at::Tensor> & save_mean, const ::std::optional<at::Tensor> & save_var_transform, bool train, double eps, ::std::array<bool,3> output_mask, const at::Tensor & reservedSpace) {
+    return at::_ops::_batch_norm_impl_index_backward::call(impl_index, input, grad_output, weight, running_mean, running_var, save_mean, save_var_transform, train, eps, output_mask, reservedSpace);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_cslt_compress.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/_cslt_compress_ops.h>
+
+namespace at {
+
+
+// aten::_cslt_compress(Tensor input) -> Tensor
+inline at::Tensor _cslt_compress(const at::Tensor & input) {
+    return at::_ops::_cslt_compress::call(input);
+}
+
+}

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

@@ -0,0 +1,35 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_addcdiv_scalar_slow(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value=1);
+TORCH_API void _foreach_addcdiv_Scalar_out(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value, at::TensorList out);
+TORCH_API void foreach_tensor_addcdiv_scalar_slow_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value=1);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_addcdiv_scalar_cuda(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value=1);
+TORCH_API void foreach_tensor_addcdiv_scalar_cuda_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value=1);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_addcdiv_scalarlist_slow(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars);
+TORCH_API void _foreach_addcdiv_ScalarList_out(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars, at::TensorList out);
+TORCH_API void foreach_tensor_addcdiv_scalarlist_slow_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_addcdiv_scalarlist_cuda(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars);
+TORCH_API void foreach_tensor_addcdiv_scalarlist_cuda_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_addcdiv_tensor_slow(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars);
+TORCH_API void _foreach_addcdiv_Tensor_out(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars, at::TensorList out);
+TORCH_API void foreach_tensor_addcdiv_tensor_slow_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_addcdiv_tensor_cuda(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars);
+TORCH_API void foreach_tensor_addcdiv_tensor_cuda_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,38 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API ::std::vector<at::Tensor> _foreach_div(at::TensorList self, const at::Scalar & scalar);
+TORCH_API void _foreach_div_out(at::TensorList out, at::TensorList self, const at::Scalar & scalar);
+TORCH_API void _foreach_div_outf(at::TensorList self, const at::Scalar & scalar, at::TensorList out);
+TORCH_API void _foreach_div_(at::TensorList self, const at::Scalar & scalar);
+TORCH_API ::std::vector<at::Tensor> _foreach_div(at::TensorList self, at::TensorList other);
+TORCH_API void _foreach_div_out(at::TensorList out, at::TensorList self, at::TensorList other);
+TORCH_API void _foreach_div_outf(at::TensorList self, at::TensorList other, at::TensorList out);
+TORCH_API void _foreach_div_(at::TensorList self, at::TensorList other);
+TORCH_API ::std::vector<at::Tensor> _foreach_div(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+TORCH_API void _foreach_div_out(at::TensorList out, at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+TORCH_API void _foreach_div_outf(at::TensorList self, at::ArrayRef<at::Scalar> scalars, at::TensorList out);
+TORCH_API void _foreach_div_(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+TORCH_API ::std::vector<at::Tensor> _foreach_div(at::TensorList self, const at::Tensor & other);
+TORCH_API void _foreach_div_out(at::TensorList out, at::TensorList self, const at::Tensor & other);
+TORCH_API void _foreach_div_outf(at::TensorList self, const at::Tensor & other, at::TensorList out);
+TORCH_API void _foreach_div_(at::TensorList self, const at::Tensor & other);
+
+} // namespace compositeexplicitautograd
+} // namespace at

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_sub_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::vector<at::Tensor> _foreach_sub(at::TensorList self, const at::Scalar & scalar);
+TORCH_API void _foreach_sub_(at::TensorList self, const at::Scalar & scalar);
+TORCH_API ::std::vector<at::Tensor> _foreach_sub(at::TensorList self, at::TensorList other, const at::Scalar & alpha=1);
+TORCH_API void _foreach_sub_(at::TensorList self, at::TensorList other, const at::Scalar & alpha=1);
+TORCH_API ::std::vector<at::Tensor> _foreach_sub(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+TORCH_API void _foreach_sub_(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+
+} // namespace cuda
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_fused_moving_avg_obs_fq_helper_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> _fused_moving_avg_obs_fq_helper(const at::Tensor & self, const at::Tensor & observer_on, const at::Tensor & fake_quant_on, at::Tensor & running_min, at::Tensor & running_max, at::Tensor & scale, at::Tensor & zero_point, double averaging_const, int64_t quant_min, int64_t quant_max, int64_t ch_axis, bool per_row_fake_quant=false, bool symmetric_quant=false);
+
+} // namespace cuda
+} // namespace at

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_make_dep_token_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 _make_dep_token_cpu(::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={}, ::std::optional<at::MemoryFormat> memory_format=::std::nullopt);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_mkldnn_transpose_meta_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 meta {
+
+TORCH_API at::Tensor & _mkldnn_transpose_(at::Tensor & self, int64_t dim0, int64_t dim1);
+
+} // namespace meta
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_native_multi_head_attention_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> _native_multi_head_attention(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, int64_t embed_dim, int64_t num_head, const at::Tensor & qkv_weight, const at::Tensor & qkv_bias, const at::Tensor & proj_weight, const at::Tensor & proj_bias, const ::std::optional<at::Tensor> & mask={}, bool need_weights=true, bool average_attn_weights=true, ::std::optional<int64_t> mask_type=::std::nullopt);
+
+} // namespace cpu
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_scaled_dot_product_fused_attention_overrideable_backward.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/_scaled_dot_product_fused_attention_overrideable_backward_ops.h>
+
+namespace at {
+
+
+// aten::_scaled_dot_product_fused_attention_overrideable_backward(Tensor grad_out, Tensor query, Tensor key, Tensor value, Tensor attn_bias, bool[4] grad_input_mask, 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) -> (Tensor grad_query, Tensor grad_key, Tensor grad_value, Tensor grad_attn_bias)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_fused_attention_overrideable_backward(const at::Tensor & grad_out, const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const at::Tensor & attn_bias, ::std::array<bool,4> grad_input_mask, const at::Tensor & out, const at::Tensor & logsumexp, const at::Tensor & cum_seq_q, const at::Tensor & cum_seq_k, int64_t max_q, int64_t max_k, double dropout_p, bool is_causal, const at::Tensor & philox_seed, const at::Tensor & philox_offset, ::std::optional<double> scale=::std::nullopt) {
+    return at::_ops::_scaled_dot_product_fused_attention_overrideable_backward::call(grad_out, query, key, value, attn_bias, grad_input_mask, out, logsumexp, cum_seq_q, cum_seq_k, max_q, max_k, dropout_p, is_causal, philox_seed, philox_offset, scale);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_fused_attention_overrideable_backward(const at::Tensor & grad_out, const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const at::Tensor & attn_bias, ::std::array<bool,4> grad_input_mask, const at::Tensor & out, const at::Tensor & logsumexp, const at::Tensor & cum_seq_q, const at::Tensor & cum_seq_k, int64_t max_q, int64_t max_k, double dropout_p, bool is_causal, const at::Tensor & philox_seed, const at::Tensor & philox_offset, ::std::optional<double> scale=::std::nullopt) {
+    return at::_ops::_scaled_dot_product_fused_attention_overrideable_backward::call(grad_out, query, key, value, attn_bias, grad_input_mask, out, logsumexp, cum_seq_q, cum_seq_k, max_q, max_k, dropout_p, is_causal, philox_seed, philox_offset, scale);
+  }
+}
+
+// aten::_scaled_dot_product_fused_attention_overrideable_backward(Tensor grad_out, Tensor query, Tensor key, Tensor value, Tensor attn_bias, bool[4] grad_input_mask, 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) -> (Tensor grad_query, Tensor grad_key, Tensor grad_value, Tensor grad_attn_bias)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_fused_attention_overrideable_backward_symint(const at::Tensor & grad_out, const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const at::Tensor & attn_bias, ::std::array<bool,4> grad_input_mask, 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::nullopt) {
+    return at::_ops::_scaled_dot_product_fused_attention_overrideable_backward::call(grad_out, query, key, value, attn_bias, grad_input_mask, out, logsumexp, cum_seq_q, cum_seq_k, max_q, max_k, dropout_p, is_causal, philox_seed, philox_offset, scale);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> _scaled_dot_product_fused_attention_overrideable_backward(const at::Tensor & grad_out, const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const at::Tensor & attn_bias, ::std::array<bool,4> grad_input_mask, 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::nullopt) {
+    return at::_ops::_scaled_dot_product_fused_attention_overrideable_backward::call(grad_out, query, key, value, attn_bias, grad_input_mask, out, logsumexp, cum_seq_q, cum_seq_k, max_q, max_k, dropout_p, is_causal, philox_seed, philox_offset, scale);
+  }
+}
+
+}

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_softmax_backward_data_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 _softmax_backward_data(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype);
+TORCH_API at::Tensor & _softmax_backward_data_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype);
+TORCH_API at::Tensor & _softmax_backward_data_outf(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype, at::Tensor & grad_input);
+
+} // namespace cuda
+} // namespace at

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

@@ -0,0 +1,24 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor & _to_dense_out(const at::Tensor & self, ::std::optional<at::ScalarType> dtype, ::std::optional<bool> masked_grad, at::Tensor & out);
+TORCH_API at::Tensor sparse_to_dense(const at::Tensor & self, ::std::optional<at::ScalarType> dtype=::std::nullopt, ::std::optional<bool> masked_grad=::std::nullopt);
+TORCH_API at::Tensor sparse_compressed_to_dense(const at::Tensor & self, ::std::optional<at::ScalarType> dtype=::std::nullopt, ::std::optional<bool> masked_grad=::std::nullopt);
+TORCH_API at::Tensor mkldnn_to_dense(const at::Tensor & self, ::std::optional<at::ScalarType> dtype=::std::nullopt, ::std::optional<bool> masked_grad=::std::nullopt);
+} // namespace native
+} // namespace at

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/adaptive_max_pool2d_backward_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/adaptive_max_pool2d_backward_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_adaptive_max_pool2d_backward_out_cpu : public at::meta::structured_adaptive_max_pool2d_backward {
+void impl(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices, const at::Tensor & grad_input);
+};
+struct TORCH_API structured_adaptive_max_pool2d_backward_out_cuda : public at::meta::structured_adaptive_max_pool2d_backward {
+void impl(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & indices, const at::Tensor & grad_input);
+};
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/adaptive_max_pool2d_meta_dispatch.h

@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> adaptive_max_pool2d(const at::Tensor & self, at::IntArrayRef output_size);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> adaptive_max_pool2d_out(at::Tensor & out, at::Tensor & indices, const at::Tensor & self, at::IntArrayRef output_size);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> adaptive_max_pool2d_outf(const at::Tensor & self, at::IntArrayRef output_size, at::Tensor & out, at::Tensor & indices);
+
+} // namespace meta
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/add.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/add_ops.h>
+
+namespace at {
+
+
+// aten::add.Tensor(Tensor self, Tensor other, *, Scalar alpha=1) -> Tensor
+inline at::Tensor add(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1) {
+    return at::_ops::add_Tensor::call(self, other, alpha);
+}
+
+// aten::add.out(Tensor self, Tensor other, *, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & add_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1) {
+    return at::_ops::add_out::call(self, other, alpha, out);
+}
+// aten::add.out(Tensor self, Tensor other, *, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & add_outf(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out) {
+    return at::_ops::add_out::call(self, other, alpha, out);
+}
+
+// aten::add.Scalar(Tensor self, Scalar other, Scalar alpha=1) -> Tensor
+inline at::Tensor add(const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha=1) {
+    return at::_ops::add_Scalar::call(self, other, alpha);
+}
+
+// aten::add.Scalar_out(Tensor self, Scalar other, Scalar alpha=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & add_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha=1) {
+    return at::_ops::add_Scalar_out::call(self, other, alpha, out);
+}
+// aten::add.Scalar_out(Tensor self, Scalar other, Scalar alpha=1, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & add_outf(const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha, at::Tensor & out) {
+    return at::_ops::add_Scalar_out::call(self, other, alpha, out);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/addbmm.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/addbmm_ops.h>
+
+namespace at {
+
+
+// aten::addbmm.out(Tensor self, Tensor batch1, Tensor batch2, *, Scalar beta=1, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & addbmm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & batch1, const at::Tensor & batch2, const at::Scalar & beta=1, const at::Scalar & alpha=1) {
+    return at::_ops::addbmm_out::call(self, batch1, batch2, beta, alpha, out);
+}
+// aten::addbmm.out(Tensor self, Tensor batch1, Tensor batch2, *, Scalar beta=1, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & addbmm_outf(const at::Tensor & self, const at::Tensor & batch1, const at::Tensor & batch2, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out) {
+    return at::_ops::addbmm_out::call(self, batch1, batch2, beta, alpha, out);
+}
+
+// aten::addbmm(Tensor self, Tensor batch1, Tensor batch2, *, Scalar beta=1, Scalar alpha=1) -> Tensor
+inline at::Tensor addbmm(const at::Tensor & self, const at::Tensor & batch1, const at::Tensor & batch2, const at::Scalar & beta=1, const at::Scalar & alpha=1) {
+    return at::_ops::addbmm::call(self, batch1, batch2, beta, alpha);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/adjoint.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/adjoint_ops.h>
+
+namespace at {
+
+
+// aten::adjoint(Tensor(a) self) -> Tensor(a)
+inline at::Tensor adjoint(const at::Tensor & self) {
+    return at::_ops::adjoint::call(self);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/arctan2_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 arctan2(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & arctan2_out(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & arctan2_(at::Tensor & self, const at::Tensor & other);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/as_strided_cpu_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 cpu {
+
+TORCH_API at::Tensor as_strided(const at::Tensor & self, at::IntArrayRef size, at::IntArrayRef stride, ::std::optional<int64_t> storage_offset=::std::nullopt);
+TORCH_API at::Tensor as_strided_symint(const at::Tensor & self, c10::SymIntArrayRef size, c10::SymIntArrayRef stride, ::std::optional<c10::SymInt> storage_offset=::std::nullopt);
+
+} // namespace cpu
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/bitwise_not_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>
+#include <ATen/ops/bitwise_not_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_bitwise_not_out : public at::meta::structured_bitwise_not {
+void impl(const at::Tensor & self, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,105 @@
+#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 bitwise_or_Tensor_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::bitwise_or")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Tensor_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "bitwise_or.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+};
+
+struct TORCH_API bitwise_or_Scalar_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::bitwise_or")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Scalar_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "bitwise_or.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+};
+
+struct TORCH_API bitwise_or_Scalar {
+  using schema = at::Tensor (const at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::bitwise_or")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Scalar")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "bitwise_or.Scalar(Tensor self, Scalar other) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Scalar & other);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Scalar & other);
+};
+
+struct TORCH_API bitwise_or_Scalar_Tensor {
+  using schema = at::Tensor (const at::Scalar &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::bitwise_or")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Scalar_Tensor")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "bitwise_or.Scalar_Tensor(Scalar self, Tensor other) -> Tensor")
+  static at::Tensor call(const at::Scalar & self, const at::Tensor & other);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & self, const at::Tensor & other);
+};
+
+struct TORCH_API bitwise_or_Tensor {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::bitwise_or")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Tensor")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "bitwise_or.Tensor(Tensor self, Tensor other) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & other);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other);
+};
+
+struct TORCH_API bitwise_or__Scalar {
+  using schema = at::Tensor & (at::Tensor &, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::bitwise_or_")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Scalar")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "bitwise_or_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!)")
+  static at::Tensor & call(at::Tensor & self, const at::Scalar & other);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Scalar & other);
+};
+
+struct TORCH_API bitwise_or__Tensor {
+  using schema = at::Tensor & (at::Tensor &, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::bitwise_or_")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Tensor")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "bitwise_or_.Tensor(Tensor(a!) self, Tensor other) -> Tensor(a!)")
+  static at::Tensor & call(at::Tensor & self, const at::Tensor & other);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, at::Tensor & self, const at::Tensor & other);
+};
+
+struct TORCH_API bitwise_or_Scalar_Tensor_out {
+  using schema = at::Tensor & (const at::Scalar &, const at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::bitwise_or")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Scalar_Tensor_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "bitwise_or.Scalar_Tensor_out(Scalar self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Scalar & self, const at::Tensor & other, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & self, const at::Tensor & other, at::Tensor & out);
+};
+
+}} // namespace at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/cholesky_solve_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 cholesky_solve_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::cholesky_solve")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "cholesky_solve.out(Tensor self, Tensor input2, bool upper=False, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & input2, bool upper, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & input2, bool upper, at::Tensor & out);
+};
+
+struct TORCH_API cholesky_solve {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::cholesky_solve")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "cholesky_solve(Tensor self, Tensor input2, bool upper=False) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & input2, bool upper);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & input2, bool upper);
+};
+
+}} // namespace at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/col_indices_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 col_indices_default(const at::Tensor & self);
+TORCH_API at::Tensor col_indices_sparse_csr(const at::Tensor & self);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API convolution_backward {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, at::OptionalSymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, c10::SymIntArrayRef, c10::SymInt, ::std::array<bool,3>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::convolution_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "convolution_backward(Tensor grad_output, Tensor input, Tensor weight, SymInt[]? bias_sizes, SymInt[] stride, SymInt[] padding, SymInt[] dilation, bool transposed, SymInt[] output_padding, SymInt groups, bool[3] output_mask) -> (Tensor, Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> call(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);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, 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);
+};
+
+struct TORCH_API convolution_backward_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> (const at::Tensor &, const at::Tensor &, const at::Tensor &, at::OptionalSymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, c10::SymIntArrayRef, c10::SymInt, ::std::array<bool,3>, at::Tensor &, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::convolution_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "convolution_backward.out(Tensor grad_output, Tensor input, Tensor weight, SymInt[]? bias_sizes, SymInt[] stride, SymInt[] padding, SymInt[] dilation, bool transposed, SymInt[] output_padding, SymInt groups, bool[3] output_mask, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))")
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> call(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, 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);
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, 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 at::_ops

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/eye_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>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor eye(int64_t n, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor & eye_out_cpu(int64_t n, at::Tensor & out);
+TORCH_API at::Tensor & eye_out_cuda(int64_t n, at::Tensor & out);
+TORCH_API at::Tensor eye(int64_t n, int64_t m, ::std::optional<at::ScalarType> dtype={}, ::std::optional<at::Layout> layout={}, ::std::optional<at::Device> device={}, ::std::optional<bool> pin_memory={});
+TORCH_API at::Tensor & eye_out_cpu(int64_t n, int64_t m, at::Tensor & out);
+TORCH_API at::Tensor & eye_out_cuda(int64_t n, int64_t m, at::Tensor & out);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,91 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+
+
+
+#include <ATen/ops/fft_rfft_ops.h>
+
+namespace at {
+
+
+// aten::fft_rfft(Tensor self, SymInt? n=None, int dim=-1, str? norm=None) -> Tensor
+inline at::Tensor fft_rfft(const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_rfft::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor fft_rfft(const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_rfft::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm);
+  }
+}
+
+// aten::fft_rfft(Tensor self, SymInt? n=None, int dim=-1, str? norm=None) -> Tensor
+inline at::Tensor fft_rfft_symint(const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_rfft::call(self, n, dim, norm);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor fft_rfft(const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_rfft::call(self, n, dim, norm);
+  }
+}
+
+// aten::fft_rfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_rfft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_rfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & fft_rfft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<int64_t> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_rfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_rfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_rfft_outf(const at::Tensor & self, ::std::optional<int64_t> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_rfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & fft_rfft_outf(const at::Tensor & self, ::std::optional<int64_t> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_rfft_out::call(self, n.has_value() ? ::std::make_optional(c10::SymInt(*n)) : ::std::nullopt, dim, norm, out);
+  }
+}
+
+// aten::fft_rfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_rfft_symint_out(at::Tensor & out, const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_rfft_out::call(self, n, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & fft_rfft_out(at::Tensor & out, const at::Tensor & self, ::std::optional<c10::SymInt> n=::std::nullopt, int64_t dim=-1, ::std::optional<c10::string_view> norm=::std::nullopt) {
+    return at::_ops::fft_rfft_out::call(self, n, dim, norm, out);
+  }
+}
+
+// aten::fft_rfft.out(Tensor self, SymInt? n=None, int dim=-1, str? norm=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fft_rfft_symint_outf(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_rfft_out::call(self, n, dim, norm, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & fft_rfft_outf(const at::Tensor & self, ::std::optional<c10::SymInt> n, int64_t dim, ::std::optional<c10::string_view> norm, at::Tensor & out) {
+    return at::_ops::fft_rfft_out::call(self, n, dim, norm, out);
+  }
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fmin_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 fmin(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & fmin_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & fmin_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/frexp_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 ::std::tuple<at::Tensor &,at::Tensor &> frexp_out(at::Tensor & mantissa, at::Tensor & exponent, const at::Tensor & self);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> frexp_outf(const at::Tensor & self, at::Tensor & mantissa, at::Tensor & exponent);
+
+} // namespace cuda
+} // namespace at

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

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

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

@@ -0,0 +1,22 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <optional>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> gru(const at::Tensor & input, const at::Tensor & hx, at::TensorList params, bool has_biases, int64_t num_layers, double dropout, bool train, bool bidirectional, bool batch_first);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> gru(const at::Tensor & data, const at::Tensor & batch_sizes, const at::Tensor & hx, at::TensorList params, bool has_biases, int64_t num_layers, double dropout, bool train, bool bidirectional);
+} // namespace native
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/hardtanh_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 hardtanh(const at::Tensor & self, const at::Scalar & min_val=-1, const at::Scalar & max_val=1);
+TORCH_API at::Tensor & hardtanh_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & min_val=-1, const at::Scalar & max_val=1);
+TORCH_API at::Tensor & hardtanh_outf(const at::Tensor & self, const at::Scalar & min_val, const at::Scalar & max_val, at::Tensor & out);
+TORCH_API at::Tensor & hardtanh_(at::Tensor & self, const at::Scalar & min_val=-1, const at::Scalar & max_val=1);
+
+} // namespace cpu
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/huber_loss_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 huber_loss(const at::Tensor & self, const at::Tensor & target, int64_t reduction=at::Reduction::Mean, double delta=1.0);
+TORCH_API at::Tensor & huber_loss_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & target, int64_t reduction=at::Reduction::Mean, double delta=1.0);
+TORCH_API at::Tensor & huber_loss_outf(const at::Tensor & self, const at::Tensor & target, int64_t reduction, double delta, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/index_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 index(const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices);
+TORCH_API at::Tensor & index_out(at::Tensor & out, const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices);
+TORCH_API at::Tensor & index_outf(const at::Tensor & self, const c10::List<::std::optional<at::Tensor>> & indices, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/isposinf_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 isposinf(const at::Tensor & self);
+TORCH_API at::Tensor & isposinf_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & isposinf_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_inv.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_inv_ops.h>
+
+namespace at {
+
+
+// aten::linalg_inv(Tensor A) -> Tensor
+inline at::Tensor linalg_inv(const at::Tensor & A) {
+    return at::_ops::linalg_inv::call(A);
+}
+
+// aten::linalg_inv.out(Tensor A, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_inv_out(at::Tensor & out, const at::Tensor & A) {
+    return at::_ops::linalg_inv_out::call(A, out);
+}
+// aten::linalg_inv.out(Tensor A, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_inv_outf(const at::Tensor & A, at::Tensor & out) {
+    return at::_ops::linalg_inv_out::call(A, out);
+}
+
+}

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_svdvals_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_svdvals(const at::Tensor & A, ::std::optional<c10::string_view> driver=::std::nullopt);
+TORCH_API at::Tensor & linalg_svdvals_out(at::Tensor & out, const at::Tensor & A, ::std::optional<c10::string_view> driver=::std::nullopt);
+TORCH_API at::Tensor & linalg_svdvals_outf(const at::Tensor & A, ::std::optional<c10::string_view> driver, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/log_softmax_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 log_softmax(const at::Tensor & self, int64_t dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+TORCH_API at::Tensor log_softmax(const at::Tensor & self, at::Dimname dim, ::std::optional<at::ScalarType> dtype=::std::nullopt);
+
+} // namespace compositeimplicitautograd
+} // namespace at

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/logspace_cpu_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 cpu {
+
+TORCH_API at::Tensor & logspace_out(at::Tensor & out, const at::Scalar & start, const at::Scalar & end, int64_t steps, double base=10.0);
+TORCH_API at::Tensor & logspace_outf(const at::Scalar & start, const at::Scalar & end, int64_t steps, double base, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mT_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 mT {
+  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::mT")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "mT(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/margin_ranking_loss.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/margin_ranking_loss_ops.h>
+
+namespace at {
+
+
+// aten::margin_ranking_loss(Tensor input1, Tensor input2, Tensor target, float margin=0.0, int reduction=Mean) -> Tensor
+inline at::Tensor margin_ranking_loss(const at::Tensor & input1, const at::Tensor & input2, const at::Tensor & target, double margin=0.0, int64_t reduction=at::Reduction::Mean) {
+    return at::_ops::margin_ranking_loss::call(input1, input2, target, margin, reduction);
+}
+
+}

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

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

.venv/lib/python3.11/site-packages/torch/include/ATen/ops/max_pool1d_with_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 max_pool1d_with_indices {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, at::IntArrayRef, at::IntArrayRef, at::IntArrayRef, at::IntArrayRef, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::max_pool1d_with_indices")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "max_pool1d_with_indices(Tensor self, int[1] kernel_size, int[1] stride=[], int[1] padding=0, int[1] dilation=1, bool ceil_mode=False) -> (Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool ceil_mode);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool ceil_mode);
+};
+
+}} // namespace at::_ops