abs() | |
abs_() | |
absolute() | |
acos() | |
acos_() | |
acosh() | |
acosh_() | |
adaptive_avg_pool1d() | |
_adaptive_avg_pool2d() | |
_adaptive_avg_pool3d() | |
adaptive_max_pool1d() | |
add() | |
_add_batch_dim() | |
addbmm() | |
addcdiv() | |
addcmul() | |
addmm() | |
_addmm_activation() | |
addmv() | |
addmv_() | |
addr() | |
_add_relu() | |
_add_relu_() | |
adjoint() | |
affine_grid_generator() | |
AggregationType | ? |
alias_copy() | |
AliasDb | ? |
align_tensors() | |
all() | |
allclose() | |
alpha_dropout() | |
alpha_dropout_() | |
amax() | |
amin() | |
_aminmax() | |
aminmax() | |
amp | Module |
_amp_foreach_non_finite_check_and_unscale_() | |
_amp_update_scale_() | |
angle() | |
Any | ? |
any() | |
AnyType | ? |
ao | Module |
arange() | |
arccos() | |
arccos_() | |
arccosh() | |
arccosh_() | |
arcsin() | |
arcsin_() | |
arcsinh() | |
arcsinh_() | |
arctan() | |
arctan_() | |
arctan2() | |
arctanh() | |
arctanh_() | |
are_deterministic_algorithms_enabled() | |
argmax() | |
argmin() | |
argsort() | |
Argument | ? |
ArgumentSpec | ? |
argwhere() | |
asarray() | |
asin() | |
asin_() | |
asinh() | |
asinh_() | |
_assert() | |
_assert_async() | |
_assert_tensor_metadata() | |
as_strided() | |
as_strided_() | |
as_strided_copy() | |
as_strided_scatter() | |
as_tensor() | |
atan() | |
atan_() | |
atan2() | |
atanh() | |
atanh_() | |
atleast_1d() | |
atleast_2d() | |
atleast_3d() | |
attr | str object |
autocast | torch.amp.autocast_mode.autocast class |
autocast_decrement_nesting() | |
autocast_increment_nesting() | |
autograd | Module |
AVG | ? |
avg_pool1d() | |
_awaits | Module |
AwaitType | ? |
backends | Module |
baddbmm() | |
bartlett_window() | |
batch_norm() | |
batch_norm_backward_elemt() | |
batch_norm_backward_reduce() | |
batch_norm_elemt() | |
batch_norm_gather_stats() | |
batch_norm_gather_stats_with_counts() | |
_batch_norm_impl_index() | |
batch_norm_stats() | |
batch_norm_update_stats() | |
BenchmarkConfig | ? |
BenchmarkExecutionStats | ? |
bernoulli() | |
bfloat16 | ? |
BFloat16Storage | ? |
BFloat16Tensor | ? |
bilinear() | |
binary_cross_entropy_with_logits() | |
bincount() | |
binomial() | |
bitwise_and() | |
bitwise_left_shift() | |
bitwise_not() | |
bitwise_or() | |
bitwise_right_shift() | |
bitwise_xor() | |
blackman_window() | |
Block | ? |
block_diag() | |
bmm() | Batch matrix-matrix product. Compare mm() |
bool | ? |
BoolStorage | ? |
BoolTensor | ? |
BoolType | ? |
broadcast_shapes() | |
broadcast_tensors() | |
broadcast_to() | |
bucketize() | |
BufferDict | ? |
builtins | Module |
ByteStorage | ? |
ByteTensor | ? |
_C | Module |
Callable | ? |
CallStack | ? |
can_cast() | |
candidate() | |
Capsule | ? |
cartesian_prod() | |
_cast_Byte() | |
_cast_Char() | |
_cast_Double() | |
_cast_Float() | |
_cast_Half() | |
_cast_Int() | |
_cast_Long() | |
_cast_Short() | |
cat() | Concatenate a sequence of tensors along a given dimension. Compare with stack() |
ccol_indices_copy() | |
cdist() | |
cdouble | ? |
ceil() | |
ceil_() | |
celu() | |
celu_() | |
cfloat | ? |
chain_matmul() | |
chalf | ? |
channel_shuffle() | |
channels_last | ? |
channels_last_3d | ? |
CharStorage | ? |
CharTensor | ? |
cholesky() | |
cholesky_inverse() | |
cholesky_solve() | |
choose_qparams_optimized() | |
_choose_qparams_per_tensor() | |
chunk() | |
_chunk_grad_outputs_efficient_attention() | |
clamp() | |
clamp_() | |
clamp_max() | |
clamp_max_() | |
clamp_min() | |
clamp_min_() | |
_classes | Module |
classes | Module |
classproperty() | |
ClassType | ? |
clear_autocast_cache() | |
clip() | |
clip_() | |
clone() | |
_coalesce() | |
Code | ? |
col_indices_copy() | |
column_stack() | |
combinations() | |
CompilationUnit | ? |
compile() | |
compiled_with_cxx11_abi() | |
CompleteArgumentSpec | ? |
complex() | |
complex128 | ? |
complex32 | ? |
complex64 | ? |
ComplexDoubleStorage | ? |
ComplexFloatStorage | ? |
ComplexType | ? |
_compute_linear_combination() | |
concat() | |
concatenate() | |
ConcreteModuleType | ? |
ConcreteModuleTypeBuilder | ? |
_conj() | |
conj() | |
_conj_copy() | |
_conj_physical() | |
conj_physical() | |
conj_physical_() | |
constant_pad_nd() | |
contiguous_format | ? |
conv1d() | |
conv2d() | |
conv3d() | |
_convert_indices_from_coo_to_csr() | |
_convert_indices_from_csr_to_coo() | |
_convolution() | |
convolution() | |
_convolution_mode() | |
conv_tbc() | |
conv_transpose1d() | |
conv_transpose2d() | |
conv_transpose3d() | |
_copy_from() | |
_copy_from_and_resize() | |
copysign() | |
corrcoef() | |
cos() | |
cos_() | |
cosh() | |
cosh_() | |
cosine_embedding_loss() | |
cosine_similarity() | |
count_nonzero() | |
cov() | |
cpp | Module |
cpu | Module |
cross() | |
crow_indices_copy() | |
_ctc_loss() | |
ctc_loss() | |
ctypes | Module |
cuda | Module |
cudnn_affine_grid_generator() | |
cudnn_batch_norm() | |
cudnn_convolution() | |
cudnn_convolution_add_relu() | |
cudnn_convolution_relu() | |
cudnn_convolution_transpose() | |
_cudnn_ctc_loss() | |
cudnn_grid_sampler() | |
_cudnn_init_dropout_state() | |
cudnn_is_acceptable() | |
_cudnn_rnn() | |
_cudnn_rnn_flatten_weight() | |
_cufft_clear_plan_cache() | |
_cufft_get_plan_cache_max_size() | |
_cufft_get_plan_cache_size() | |
_cufft_set_plan_cache_max_size() | |
cummax() | |
_cummax_helper() | |
cummin() | |
_cummin_helper() | |
cumprod() | |
cumsum() | |
cumulative_trapezoid() | |
_debug_has_internal_overlap() | |
_decomp | Module |
DeepCopyMemoTable | ? |
default_generator | ? |
deg2rad() | |
deg2rad_() | |
dequantize() | |
DeserializationStorageContext | ? |
det() | |
detach() | |
detach_() | |
detach_copy() | |
device | A torch.device object (cpu, cuda or mps) which represents the device on which the tensor is (or will be) allocated. |
DeviceObjType | ? |
diag() | |
diag_embed() | |
diagflat() | |
diagonal() | |
diagonal_copy() | |
diagonal_scatter() | |
Dict | ? |
DictType | ? |
diff() | |
digamma() | |
_dim_arange() | |
_dirichlet_grad() | |
_disable_functionalization() | |
DisableTorchFunction | torch._C.DisableTorchFunction class |
DisableTorchFunctionSubclass | torch._C.DisableTorchFunctionSubclass class |
_dispatch | Module |
DispatchKey | ? |
DispatchKeySet | ? |
dist() | |
distributed | Module |
distributions | Module |
div() | |
divide() | |
dot() | |
double | ? |
DoubleStorage | ? |
DoubleTensor | ? |
dropout() | |
dropout_() | |
dsmm() | |
dsplit() | |
dstack() | |
dtype | torch.dtype class |
e | float object |
_efficientzerotensor() | |
eig() | |
einsum() | |
embedding() | |
_embedding_bag() | |
embedding_bag() | |
_embedding_bag_forward_only() | |
embedding_renorm_() | |
empty() | |
_empty_affine_quantized() | |
empty_like() | |
_empty_per_channel_affine_quantized() | |
empty_quantized() | |
empty_strided() | |
_enable_functionalization() | |
enable_grad | torch.autograd.grad_mode.enable_grad class. Compare with no_grad , set_grad_enabled and inference_mode |
EnumType | ? |
eq() | |
equal() | |
erf() | |
erf_() | |
erfc() | |
erfc_() | |
erfinv() | |
ErrorReport | ? |
_euclidean_dist() | |
ExcludeDispatchKeyGuard | ? |
ExecutionPlan | ? |
exp() | |
exp_() | |
exp2() | |
exp2_() | |
expand_copy() | |
expm1() | |
expm1_() | |
eye() | |
_fake_quantize_learnable_per_channel_affine() | |
_fake_quantize_learnable_per_tensor_affine() | |
fake_quantize_per_channel_affine() | |
fake_quantize_per_tensor_affine() | |
_fake_quantize_per_tensor_affine_cachemask_tensor_qparams() | |
FatalError | torch.FatalError class |
fbgemm_linear_fp16_weight() | |
fbgemm_linear_fp16_weight_fp32_activation() | |
fbgemm_linear_int8_weight() | |
fbgemm_linear_int8_weight_fp32_activation() | |
fbgemm_linear_quantize_weight() | |
fbgemm_pack_gemm_matrix_fp16() | |
fbgemm_pack_quantized_matrix() | |
feature_alpha_dropout() | |
feature_alpha_dropout_() | |
feature_dropout() | |
feature_dropout_() | |
fft | Module |
_fft_c2c() | |
_fft_c2r() | |
_fft_r2c() | |
FileCheck | ? |
fill() | |
fill_() | |
finfo | torch.finfo class |
fix() | |
fix_() | |
flatten() | |
flip() | |
fliplr() | |
flipud() | |
float | ? |
float16 | ? |
float32 | ? |
float64 | ? |
float_power() | |
FloatStorage | ? |
FloatTensor | ? |
FloatType | ? |
floor() | |
floor_() | |
floor_divide() | |
fmax() | |
fmin() | |
fmod() | |
_foobar() | |
_foreach_abs() | |
_foreach_abs_() | |
_foreach_acos() | |
_foreach_acos_() | |
_foreach_add() | |
_foreach_add_() | |
_foreach_addcdiv() | |
_foreach_addcdiv_() | |
_foreach_addcmul() | |
_foreach_addcmul_() | |
_foreach_asin() | |
_foreach_asin_() | |
_foreach_atan() | |
_foreach_atan_() | |
_foreach_ceil() | |
_foreach_ceil_() | |
_foreach_clamp_max() | |
_foreach_clamp_max_() | |
_foreach_clamp_min() | |
_foreach_clamp_min_() | |
_foreach_cos() | |
_foreach_cos_() | |
_foreach_cosh() | |
_foreach_cosh_() | |
_foreach_div() | |
_foreach_div_() | |
_foreach_erf() | |
_foreach_erf_() | |
_foreach_erfc() | |
_foreach_erfc_() | |
_foreach_exp() | |
_foreach_exp_() | |
_foreach_expm1() | |
_foreach_expm1_() | |
_foreach_floor() | |
_foreach_floor_() | |
_foreach_frac() | |
_foreach_frac_() | |
_foreach_lerp() | |
_foreach_lerp_() | |
_foreach_lgamma() | |
_foreach_lgamma_() | |
_foreach_log() | |
_foreach_log_() | |
_foreach_log10() | |
_foreach_log10_() | |
_foreach_log1p() | |
_foreach_log1p_() | |
_foreach_log2() | |
_foreach_log2_() | |
_foreach_maximum() | |
_foreach_maximum_() | |
_foreach_minimum() | |
_foreach_minimum_() | |
_foreach_mul() | |
_foreach_mul_() | |
_foreach_neg() | |
_foreach_neg_() | |
_foreach_norm() | |
_foreach_reciprocal() | |
_foreach_reciprocal_() | |
_foreach_round() | |
_foreach_round_() | |
_foreach_sigmoid() | |
_foreach_sigmoid_() | |
_foreach_sin() | |
_foreach_sin_() | |
_foreach_sinh() | |
_foreach_sinh_() | |
_foreach_sqrt() | |
_foreach_sqrt_() | |
_foreach_sub() | |
_foreach_sub_() | |
_foreach_tan() | |
_foreach_tan_() | |
_foreach_tanh() | |
_foreach_tanh_() | |
_foreach_trunc() | |
_foreach_trunc_() | |
_foreach_zero_() | |
fork() | |
frac() | |
frac_() | |
_freeze_functional_tensor() | |
frexp() | |
frobenius_norm() | |
frombuffer() | |
from_dlpack() | |
from_file() | |
_from_functional_tensor() | |
from_numpy() | Compare with Tensor.numpy() |
full() | |
full_like() | |
func | Module |
functional | Module |
FunctionSchema | ? |
_functorch | Module |
_fused_adam_() | |
_fused_adamw_() | |
_fused_dropout() | |
fused_moving_avg_obs_fake_quant() | |
_fused_moving_avg_obs_fq_helper() | |
_fused_sdp_choice() | |
Future | ? |
futures | Module |
FutureType | ? |
_fw_primal_copy() | |
fx | Module |
gather() | |
gcd() | |
gcd_() | |
ge() | |
Generator | torch._C.Generator class |
geqrf() | |
ger() | |
get_autocast_cpu_dtype() | |
get_autocast_gpu_dtype() | |
get_default_dtype() | |
get_deterministic_debug_mode() | |
get_device() | |
get_file_path() | |
get_float32_matmul_precision() | |
get_num_interop_threads() | |
get_num_threads() | |
get_rng_state() | |
_GLOBAL_DEVICE_CONTEXT | NoneType object |
Gradient | ? |
gradient() | |
Graph | ? |
GraphExecutorState | ? |
greater() | |
greater_equal() | |
grid_sampler() | |
grid_sampler_2d() | |
_grid_sampler_2d_cpu_fallback() | |
grid_sampler_3d() | |
group_norm() | |
gru() | |
gru_cell() | |
gt() | |
_guards | Module |
half | ? |
HalfStorage | ? |
HalfTensor | ? |
hamming_window() | |
hann_window() | |
hardshrink() | |
_has_compatible_shallow_copy_type() | |
has_cuda | bool object |
has_cudnn | bool object |
has_lapack | bool object |
has_mkl | bool object |
has_mkldnn | bool object |
has_mps | bool object |
has_openmp | bool object |
has_spectral | bool object |
heaviside() | |
hinge_embedding_loss() | |
histc() | |
histogram() | |
histogramdd() | |
_histogramdd_bin_edges() | |
_histogramdd_from_bin_cts() | |
_histogramdd_from_bin_tensors() | |
hsmm() | |
hsplit() | |
hspmm() | |
hstack() | |
hub | Module |
hypot() | |
i0() | |
i0_() | |
igamma() | |
igammac() | |
iinfo | torch.iinfo class |
imag() | |
_import_dotted_name() | |
import_ir_module() | |
import_ir_module_from_buffer() | |
index_add() | |
index_copy() | |
index_fill() | |
index_put() | |
index_put_() | |
_index_put_impl_() | |
index_reduce() | |
index_select() | |
_indices_copy() | |
indices_copy() | |
inf | float object |
inference_mode | torch.autograd.grad_mode.inference_mode class. Compare with enable_grad, no_grad and set_grad_enabled`. |
InferredType | ? |
_initExtension() | |
initial_seed() | |
init_num_threads() | |
inner() | |
inspect | Module |
instance_norm() | |
int | ? |
int16 | ? |
int32 | ? |
int64 | ? |
int8 | ? |
InterfaceType | ? |
int_repr() | |
IntStorage | ? |
IntTensor | ? |
IntType | ? |
inverse() | |
IODescriptor | ? |
_is_all_true() | |
is_anomaly_check_nan_enabled() | |
is_anomaly_enabled() | |
_is_any_true() | |
is_autocast_cache_enabled() | |
is_autocast_cpu_enabled() | |
is_autocast_enabled() | |
isclose() | |
is_complex() | |
is_conj() | |
is_deterministic_algorithms_warn_only_enabled() | |
is_distributed() | |
isfinite() | |
is_floating_point() | |
_is_functional_tensor() | |
is_grad_enabled() | |
isin() | |
isinf() | |
is_inference() | |
is_inference_mode_enabled() | |
isnan() | |
is_neg() | |
isneginf() | |
is_nonzero() | |
isposinf() | |
isreal() | |
is_same_size() | |
is_signed() | |
is_storage() | |
is_tensor() | |
istft() | |
is_vulkan_available() | |
is_warn_always_enabled() | |
_is_zerotensor() | |
jit | Module |
JITException | torch.jit.Error class |
_jit_internal | Module |
kaiser_window() | |
kl_div() | |
kron() | |
kthvalue() | |
layer_norm() | |
layout | torch.layout class |
lcm() | |
lcm_() | |
ldexp() | |
ldexp_() | |
le() | |
legacy_contiguous_format | ? |
lerp() | |
less() | |
less_equal() | |
lgamma() | |
library | Module |
linalg | Module |
_linalg_check_errors() | |
_linalg_det() | |
_linalg_eigh() | |
_linalg_slogdet() | |
_linalg_solve_ex() | |
_linalg_svd() | |
_linalg_utils | Module |
linspace() | |
ListType | ? |
LiteScriptModule | ? |
load() | load an object from a file that was saved with torch.save() |
_load_global_deps() | |
_lobpcg | Module |
lobpcg() | |
LockingLogger | ? |
log() | |
log_() | |
log10() | |
log10_() | |
log1p() | |
log1p_() | |
log2() | |
log2_() | |
logaddexp() | |
logaddexp2() | |
_logcumsumexp() | |
logcumsumexp() | |
logdet() | |
LoggerBase | ? |
logical_and() | |
logical_not() | |
logical_or() | |
logical_xor() | |
logit() | |
logit_() | |
_log_softmax() | |
log_softmax() | |
_log_softmax_backward_data() | |
logspace() | |
logsumexp() | |
long | ? |
LongStorage | ? |
LongTensor | ? |
_lowrank | Module |
lstm() | |
lstm_cell() | |
_lstm_mps() | |
lstsq() | |
lt() | |
lu() | |
lu_solve() | |
lu_unpack() | |
_lu_with_info() | |
_make_dual() | |
_make_dual_copy() | |
_make_per_channel_quantized_tensor() | |
_make_per_tensor_quantized_tensor() | |
manual_seed() | Seeds the random number generator and returns a torch.Generator object. See also David Picard: torch.manual_seed(3407) is all you need: On the influence of random seeds in deep learning architectures for computer vision. |
margin_ranking_loss() | |
masked | Module |
masked_fill() | |
_masked_scale() | |
masked_scatter() | |
masked_select() | |
_masked_softmax() | |
math | Module |
matmul() | Matrix mulitplication, dot product, matrix-vector product, depending on input shapes, with broadcasting. Compare with mm() |
matrix_exp() | |
matrix_power() | |
matrix_rank() | |
max() | |
maximum() | |
max_pool1d() | |
max_pool1d_with_indices() | |
max_pool2d() | |
max_pool3d() | |
mean() | |
median() | |
memory_format | torch.memory_format class |
merge_type_from_type_comment() | |
meshgrid() | |
_meta_registrations | Module |
min() | |
minimum() | |
miopen_batch_norm() | |
miopen_convolution() | |
miopen_convolution_add_relu() | |
miopen_convolution_relu() | |
miopen_convolution_transpose() | |
miopen_depthwise_convolution() | |
miopen_rnn() | |
_mkldnn | ? |
mkldnn_adaptive_avg_pool2d() | |
mkldnn_convolution() | |
mkldnn_linear_backward_weights() | |
mkldnn_max_pool2d() | |
mkldnn_max_pool3d() | |
_mkldnn_reshape() | |
mkldnn_rnn_layer() | |
_mkldnn_transpose() | |
_mkldnn_transpose_() | |
mm() | Matrix mulitplication (no broadcasting). Compare with matmul() , smm() and bmm() |
mode() | |
ModuleDict | ? |
moveaxis() | |
movedim() | |
_mps_convolution() | |
_mps_convolution_transpose() | |
msort() | |
mul() | |
multinomial() | |
multiply() | |
multiprocessing | Module |
mv() | |
mvlgamma() | |
name | str object |
_namedtensor_internals | Module |
nan | float object |
nanmean() | |
nanmedian() | |
nanquantile() | |
nansum() | |
nan_to_num() | |
nan_to_num_() | |
narrow() | |
narrow_copy() | |
native_batch_norm() | |
_native_batch_norm_legit() | |
native_channel_shuffle() | |
_native_decoder_only_multi_head_attention() | |
native_dropout() | |
native_group_norm() | |
native_layer_norm() | |
_native_multi_head_attention() | |
native_norm() | |
ne() | |
neg() | |
neg_() | |
negative() | |
negative_() | |
_neg_view() | |
_neg_view_copy() | |
nested | Module |
_nested_from_padded() | |
_nested_from_padded_and_nested_example() | |
_nested_tensor_from_mask() | |
_nested_tensor_from_mask_left_aligned() | |
_nested_tensor_from_tensor_list() | |
_nested_tensor_softmax_with_shape() | |
nextafter() | |
nn | A module, which contains, for example torch.nn.Parameter , torch.nn.Linear , torch.nn.MSELoss or torch.nn.Module |
_nnpack_available() | |
_nnpack_spatial_convolution() | |
Node | ? |
no_grad | Compare with using torch.no_grad() as a decorator. see also with enable_grad, set_grad_enabled and inference_mode` |
NoneType | ? |
nonzero() | |
NoopLogger | ? |
norm() | |
normal() | |
norm_except_dim() | |
not_equal() | |
nuclear_norm() | |
NumberType | ? |
numel() | |
obj() | |
ones() | Create a tensor with a given shape, filled with the value 1 . Compare with zeros() |
ones_like() | |
onnx | Module |
OperatorInfo | ? |
_ops | Module |
ops | Module |
optim | Module |
Optional | ? |
OptionalType | ? |
orgqr() | |
ormqr() | |
os | Module |
outer() | |
overrides | Module |
package | Module |
_pack_padded_sequence() | |
_pad_packed_sequence() | |
pairwise_distance() | |
ParameterDict | ? |
parse_ir() | |
parse_schema() | |
parse_type_comment() | |
pca_lowrank() | |
pdist() | |
per_channel_affine | ? |
per_channel_affine_float_qparams | ? |
per_channel_symmetric | ? |
permute() | |
permute_copy() | |
per_tensor_affine | ? |
per_tensor_symmetric | ? |
pi | float object |
_pin_memory() | |
pinverse() | |
pixel_shuffle() | |
pixel_unshuffle() | |
platform | Module |
poisson() | |
poisson_nll_loss() | |
polar() | |
polygamma() | |
positive() | |
pow() | |
_preload_cuda_deps() | |
prelu() | |
_prelu_kernel() | |
prepare_multiprocessing_environment() | |
preserve_format | ? |
_prims | Module |
_prims_common | Module |
PRIVATE_OPS | tuple object |
prod() | |
profiler | Module |
promote_types() | |
put() | |
py_float | float class |
py_int | int class |
PyObjectType | ? |
PyTorchFileReader | ? |
PyTorchFileWriter | ? |
qint32 | ? |
QInt32Storage | ? |
qint8 | ? |
QInt8Storage | ? |
q_per_channel_axis() | |
q_per_channel_scales() | |
q_per_channel_zero_points() | |
qr() | |
q_scale() | |
qscheme | torch.qscheme class |
quantile() | |
quantization | Module |
quantized_batch_norm() | |
quantized_gru | ? |
quantized_gru_cell() | |
quantized_lstm | ? |
quantized_lstm_cell() | |
quantized_max_pool1d() | |
quantized_max_pool2d() | |
quantized_rnn_relu_cell() | |
quantized_rnn_tanh_cell() | |
quantize_per_channel() | |
quantize_per_tensor() | |
quantize_per_tensor_dynamic() | |
quasirandom | Module |
quint2x4 | ? |
QUInt2x4Storage | ? |
quint4x2 | ? |
QUInt4x2Storage | ? |
quint8 | ? |
QUInt8Storage | ? |
q_zero_point() | |
rad2deg() | |
rad2deg_() | |
random | Module |
randperm() | Random permuation of integers in range 0 … n |
range() | |
ravel() | |
read_vitals() | |
real() | |
reciprocal() | |
reciprocal_() | |
_refs | Module |
_register_device_module() | |
relu() | |
relu_() | |
remainder() | |
_remove_batch_dim() | |
renorm() | |
repeat_interleave() | |
reshape() | |
_reshape_alias_copy() | |
_reshape_from_tensor() | |
resize_as_() | |
resize_as_sparse_() | |
_resize_output_() | |
resolve_conj() | |
resolve_neg() | |
result_type() | |
return_types | Module |
rnn_relu() | |
rnn_relu_cell() | |
rnn_tanh() | |
rnn_tanh_cell() | |
roll() | |
rot90() | |
round() | |
round_() | |
row_indices_copy() | |
row_stack() | |
_rowwise_prune() | |
RRefType | ? |
rrelu() | |
rrelu_() | |
rsqrt() | |
rsqrt_() | |
rsub() | |
saddmm() | |
_sample_dirichlet() | |
_saturate_weight_to_fp16() | |
save() | Saves an object to a file (typically with a .pt , or less desirably a .pth extension) using pickle . The object can then be loaded again with torch.load() . Compare with torch.nn.Module.load_state_dict() . |
scalar_tensor() | |
_scaled_dot_product_attention_math() | |
_scaled_dot_product_efficient_attention() | |
_scaled_dot_product_flash_attention() | |
scatter() | |
scatter_add() | |
scatter_reduce() | |
ScriptClass | ? |
ScriptClassFunction | ? |
ScriptDict | ? |
ScriptDictIterator | ? |
ScriptDictKeyIterator | ? |
ScriptFunction | ? |
ScriptList | ? |
ScriptListIterator | ? |
ScriptMethod | ? |
ScriptModule | ? |
ScriptModuleSerializer | ? |
ScriptObject | ? |
ScriptObjectProperty | ? |
searchsorted() | |
seed() | |
_segment_reduce() | |
segment_reduce() | |
select() | |
select_copy() | |
select_scatter() | |
selu() | |
selu_() | |
serialization | Module |
SerializationStorageContext | ? |
Set | ? |
set_anomaly_enabled() | |
set_autocast_cache_enabled() | |
set_autocast_cpu_dtype() | |
set_autocast_cpu_enabled() | |
set_autocast_enabled() | |
set_autocast_gpu_dtype() | |
set_default_device() | |
set_default_dtype() | |
set_default_tensor_type() | |
set_deterministic_debug_mode() | |
set_float32_matmul_precision() | |
set_flush_denormal() | |
set_grad_enabled | torch.autograd.grad_mode.set_grad_enabled class. Compare with enable_grad, no_grad and inference_mode` |
set_num_interop_threads() | |
set_num_threads() | |
set_printoptions() | Set options that influence representations of data and structures when printing objects: precision, threshold, edgeitem, linewidth, profile, sci_mode. |
set_rng_state() | |
set_vital() | |
set_warn_always() | |
sgn() | |
_shape_as_tensor() | |
short | ? |
ShortStorage | ? |
ShortTensor | ? |
sigmoid() | An alias for torch.special.expit() |
sigmoid_() | |
sign() | |
signal | Module |
signbit() | |
sin() | |
sin_() | |
sinc() | |
sinc_() | |
sinh() | |
sinh_() | |
Size | torch.Size class |
slice_copy() | |
slice_scatter() | |
slogdet() | |
smm() | Matrix multiplication of a sparse and a dense matrix. Compare with mm() |
_sobol_engine_draw() | |
_sobol_engine_ff_() | |
_sobol_engine_initialize_state_() | |
_sobol_engine_scramble_() | |
_softmax() | |
softmax() | |
_softmax_backward_data() | |
solve() | |
sort() | |
_sources | Module |
sparse | Module |
_sparse_broadcast_to() | |
_sparse_broadcast_to_copy() | |
sparse_bsc | ? |
sparse_bsc_tensor() | |
sparse_bsr | ? |
sparse_bsr_tensor() | |
sparse_compressed_tensor() | |
sparse_coo | ? |
sparse_coo_tensor() | |
_sparse_coo_tensor_unsafe() | |
sparse_csc | ? |
sparse_csc_tensor() | |
sparse_csr | ? |
_sparse_csr_prod() | |
_sparse_csr_sum() | |
sparse_csr_tensor() | |
_sparse_log_softmax_backward_data() | |
_sparse_softmax_backward_data() | |
_sparse_sparse_matmul() | |
_sparse_sum() | |
special | A module which is modeled after SciPy's special module. |
split() | |
split_copy() | |
split_with_sizes() | |
split_with_sizes_copy() | |
spmm() | |
sqrt() | |
sqrt_() | |
square() | |
square_() | |
squeeze() | |
squeeze_copy() | |
sspaddmm() | |
_stack() | |
stack() | Compare with cat() |
_standard_gamma() | |
_standard_gamma_grad() | |
StaticModule | ? |
std() | |
std_mean() | |
stft() | |
Storage | ? |
storage | Module |
StorageBase | ? |
_storage_classes | set object |
Stream | torch.Stream class |
StreamObjType | ? |
strided | ? |
StringType | ? |
sub() | |
_subclasses | Module |
subtract() | |
SUM | ? |
sum() | |
svd() | |
svd_lowrank() | |
swapaxes() | |
swapdims() | |
SymBool | torch.SymBool class |
symeig() | |
SymFloat | torch.SymFloat class |
sym_float() | |
SymInt | torch.SymInt class |
sym_int() | |
SymIntType | ? |
sym_max() | |
sym_min() | |
sym_not() | |
_sync() | |
sys | Module |
t() | |
Tag | ? |
take() | |
take_along_dim() | |
tan() | |
tan_() | |
tanh() | |
tanh_() | |
t_copy() | |
Tensor | ? |
_tensor | Module |
tensor() | |
_tensor_classes | set object |
tensordot() | |
tensor_split() | |
_tensor_str | Module |
TensorType | ? |
_test_autograd_multiple_dispatch() | |
_test_autograd_multiple_dispatch_view() | |
_test_autograd_multiple_dispatch_view_copy() | |
_test_check_tensor() | |
testing | Module |
_test_serialization_subcmul() | |
textwrap | Module |
threshold() | |
threshold_() | |
ThroughputBenchmark | ? |
tile() | |
_to_cpu() | |
to_dlpack() | |
_to_functional_tensor() | |
topk() | |
torch | Module |
_TorchCompileInductorWrapper | torch._TorchCompileInductorWrapper class |
torch_version | Module |
trace() | |
TracingState | ? |
_transform_bias_rescale_qkv() | |
_transformer_decoder_only_layer_fwd() | |
_transformer_encoder_layer_fwd() | |
transpose() | |
transpose_copy() | |
trapezoid() | |
trapz() | |
triangular_solve() | |
tril() | |
tril_indices() | |
_trilinear() | |
triplet_margin_loss() | |
_triton_multi_head_attention() | |
_triton_scaled_dot_attention() | |
triu() | |
triu_indices() | |
true_divide() | |
trunc() | |
trunc_() | |
TupleType | ? |
Type | ? |
TYPE_CHECKING | bool object |
TypedStorage | torch.storage.TypedStorage class |
typename() | |
types | Module |
uint8 | ? |
unbind() | |
unbind_copy() | |
unflatten() | |
unfold_copy() | |
unify_type_list() | |
Union | ? |
UnionType | ? |
_unique() | |
unique() | |
_unique2() | |
unique_consecutive() | |
_unpack_dual() | |
unsafe_chunk() | |
unsafe_split() | |
unsafe_split_with_sizes() | |
unsqueeze() | |
unsqueeze_copy() | |
UntypedStorage | ? |
Use | ? |
_use_cudnn_ctc_loss() | |
_use_cudnn_rnn_flatten_weight() | |
use_deterministic_algorithms() | |
USE_GLOBAL_DEPS | bool object |
USE_RTLD_GLOBAL_WITH_LIBTORCH | bool object |
_utils | Module |
utils | Module |
_utils_internal | Module |
_validate_compressed_sparse_indices() | |
_validate_sparse_bsc_tensor_args() | |
_validate_sparse_bsr_tensor_args() | |
_validate_sparse_compressed_tensor_args() | |
_validate_sparse_coo_tensor_args() | |
_validate_sparse_csc_tensor_args() | |
_validate_sparse_csr_tensor_args() | |
Value | ? |
_values_copy() | |
values_copy() | |
vander() | |
var() | |
var_mean() | |
vdot() | |
version | Module |
_VF | Module |
view_as_complex() | |
view_as_complex_copy() | |
view_as_real() | |
view_as_real_copy() | |
view_copy() | |
vitals_enabled() | |
vmap() | |
_vmap_internals | Module |
vsplit() | |
vstack() | |
wait() | |
_warn_typed_storage_removal() | |
_weight_norm() | |
_weight_norm_interface() | |
_weights_only_unpickler | Module |
where() | |
windows | Module |
xlogy() | |
xlogy_() | |
zero_() | |
zeros() | Create a tensor with a given shape, filled with the value 0 . Compare with ones() |
zeros_like() | |