R Observatory
Search

torch

torch 0.0.3 to 0.17.0

CRAN

Record of 0.0.3Record of 0.17.0Both records side by side

Daily mean of CRAN downloads while each release was the newest, as blocks, and week by week as a line, one scale up to 2K. Choose a release to compare from or to it.
  • 0.0.3 and 0.17.0
  • between them
  • other releases
  • daily mean, week by week
  • after this copy’s series ends on 26 Oct 2026
05001K1.5K2K
Time between
2,031 days28 releases between
Advisories, method
None knownfor either release
Exports, method
9 removed270 added
Dependencies
15 changes11 added, 3 bounds, 1 moved11 added, 3 bounds
Licence
SameMIT + file LICENSE
Maintainer
Changednow Tomasz Kalinowski

Downloads while newest

Method, downloads

ReleaseDays newestDays countedCRAN downloadsDaily mean
0.17.0
11 Apr to 28 Oct 2026, still the newest
201199206,3861,037
0.0.3
18 to 27 Sep 2020
1025425

Daily mean while newest, 0.17.0 +3,983% up since 0.0.3

Counted on the CRAN mirror from each release's day to the day before the next release. This copy's daily series ends on 26 Oct 2026, so a span that runs past it is counted to that day and its daily mean covers the days counted. Other channels repackage on their own schedules, so they are not split by release.

  • added
  • bound or field changed
PackageField0.0.30.17.0
addedcallrImports–any version
addedcliImports–>= 3.0.0
addedcoroImports–>= 1.0.2
addeddescImports–any version
changedglueSuggests to Importsany versionany version
addedjsonliteImports–any version
addedkatexSuggests–any version
changedknitrSuggestsany version>= 1.36
addedmvtnormSuggests–any version
addednumDerivSuggests–any version
changedrlangImportsany version>= 1.0.0
addedsafetensorsImports–>= 0.1.1
addedscalesImports–any version
changedtestthatSuggests>= 2.1.0>= 3.0.0
addedtoolsImports–any version
Side by side
Measure0.0.30.17.0Change
Imports9 packages18 packages+8, 2 bounds
LinkingTo1 package1 package
Suggests6 packages8 packages+3, 3 bounds

Exports

added Added (270)

All 270
  • as_iterator
  • backends_cudnn_is_available
  • backends_cudnn_version
  • backends_mkldnn_is_available
  • backends_mkl_is_available
  • backends_mps_is_available
  • backends_openmp_is_available
  • buffer_from_torch_tensor
  • call_torch_function
  • clone_module
  • contrib_sort_vertices
  • cuda_amp_grad_scaler
  • cuda_dump_memory_snapshot
  • cuda_empty_cache
  • cuda_get_device_capability
  • cuda_get_rng_state
  • cuda_memory_snapshot
  • cuda_memory_stats
  • cuda_memory_summary
  • cuda_record_memory_history
  • cuda_runtime_version
  • cuda_set_rng_state
  • cuda_synchronize
  • dataset_subset
  • distr_bernoulli
  • distr_categorical
  • distr_chi2
  • distr_gamma
  • distr_mixture_same_family
  • distr_multivariate_normal
  • distr_normal
  • distr_poisson
  • get_install_libs_url
  • install_torch_from_file
  • install_torch_sitrep
  • is_nn_buffer
  • is_nn_module
  • is_nn_parameter
  • is_optimizer
  • iterable_dataset
  • jit_compile
  • jit_load
  • jit_ops
  • jit_save
  • jit_save_for_mobile
  • jit_scalar
  • jit_serialize
  • jit_trace
  • jit_trace_module
  • jit_tuple
  • jit_unserialize
  • linalg_cholesky
  • linalg_cholesky_ex
  • linalg_cond
  • linalg_det
  • linalg_eig
  • linalg_eigh
  • linalg_eigvals
  • linalg_eigvalsh
  • linalg_householder_product
  • linalg_inv
  • linalg_inv_ex
  • linalg_lstsq
  • linalg_matrix_norm
  • linalg_matrix_power
  • linalg_matrix_rank
  • linalg_multi_dot
  • linalg_norm
  • linalg_pinv
  • linalg_qr
  • linalg_slogdet
  • linalg_solve
  • linalg_solve_triangular
  • linalg_svd
  • linalg_svdvals
  • linalg_tensorinv
  • linalg_tensorsolve
  • linalg_vector_norm
  • local_autocast
  • local_device
  • local_enable_grad
  • local_no_grad
  • local_torch_manual_seed
  • loop
  • lr_cosine_annealing
  • lr_lambda
  • lr_multiplicative
  • lr_one_cycle
  • lr_reduce_on_plateau
  • lr_scheduler
  • lr_step
  • nnf_area_under_min_fpr_fnr
  • nnf_contrib_sparsemax
  • nnf_silu
  • nnf_triplet_margin_with_distance_loss
  • nn_aum_loss
  • nn_batch_norm3d
  • nn_bce_with_logits_loss
  • nn_buffer
  • nn_contrib_sparsemax
  • nn_cosine_embedding_loss
  • nn_ctc_loss
  • nn_embedding_bag
  • nn_flatten
  • nn_group_norm
  • nn_gru
  • nn_hinge_embedding_loss
  • nn_kl_div_loss
  • nn_l1_loss
  • nn_layer_norm
  • nn_lstm
  • nn_margin_ranking_loss
  • nn_module_dict
  • nn_mse_loss
  • nn_multilabel_margin_loss
  • nn_multilabel_soft_margin_loss
  • nn_multi_margin_loss
  • nn_nll_loss
  • nn_pairwise_distance
  • nn_parameter
  • nn_poisson_nll_loss
  • nn_prune_head
  • nn_silu
  • nn_smooth_l1_loss
  • nn_soft_margin_loss
  • nn_transformer_encoder
  • nn_transformer_encoder_layer
  • nn_triplet_margin_loss
  • nn_triplet_margin_with_distance_loss
  • nn_unflatten
  • nn_upsample
  • nn_utils_clip_grad_norm_
  • nn_utils_clip_grad_value_
  • nn_utils_weight_norm
  • optimizer
  • OptimizerIgnite
  • optimizer_ignite
  • optim_adadelta
  • optim_adagrad
  • optim_adamw
  • optim_asgd
  • optim_ignite_adagrad
  • optim_ignite_adam
  • optim_ignite_adamw
  • optim_ignite_rmsprop
  • optim_ignite_sgd
  • optim_lbfgs
  • optim_required
  • optim_rmsprop
  • optim_rprop
  • sampler
  • set_autocast
  • slc
  • torch_absolute
  • torch_acosh
  • torch_amax
  • torch_amin
  • torch_arccos
  • torch_arccosh
  • torch_arcsin
  • torch_arcsinh
  • torch_arctan
  • torch_arctanh
  • torch_asinh
  • torch_atanh
  • torch_atleast_1d
  • torch_atleast_2d
  • torch_atleast_3d
  • torch_bfloat16
  • torch_block_diag
  • torch_bucketize
  • torch_cdouble
  • torch_cfloat
  • torch_cfloat32
  • torch_cfloat64
  • torch_cfloat128
  • torch_chalf
  • torch_channel_shuffle
  • torch_clip
  • torch_clone
  • torch_complex
  • torch_count_nonzero
  • torch_deg2rad
  • torch_dequantize
  • torch_diff
  • torch_divide
  • torch_dstack
  • torch_exp2
  • torch_fft_fft
  • torch_fft_fftfreq
  • torch_fft_ifft
  • torch_fft_irfft
  • torch_fft_rfft
  • torch_fix
  • torch_fliplr
  • torch_flipud
  • torch_float8_e4m3fn
  • torch_float8_e5m2
  • torch_gcd
  • torch_get_num_interop_threads
  • torch_get_num_threads
  • torch_get_rng_state
  • torch_greater
  • torch_greater_equal
  • torch_heaviside
  • torch_hstack
  • torch_hypot
  • torch_i0
  • torch_index
  • torch_index_put
  • torch_index_put_
  • torch_install_path
  • torch_isclose
  • torch_isneginf
  • torch_isposinf
  • torch_isreal
  • torch_istft
  • torch_is_nonzero
  • torch_kaiser_window
  • torch_kron
  • torch_lcm
  • torch_ldexp
  • torch_less
  • torch_less_equal
  • torch_logaddexp
  • torch_logaddexp2
  • torch_logcumsumexp
  • torch_logit
  • torch_lu_unpack
  • torch_matrix_exp
  • torch_maximum
  • torch_minimum
  • torch_movedim
  • torch_multiply
  • torch_nanquantile
  • torch_nansum
  • torch_negative
  • torch_nextafter
  • torch_not_equal
  • torch_outer
  • torch_polar
  • torch_quantile
  • torch_rad2deg
  • torch_scalar_tensor
  • torch_scaled_dot_product_attention
  • torch_searchsorted
  • torch_serialize
  • torch_set_num_interop_threads
  • torch_set_num_threads
  • torch_set_rng_state
  • torch_sgn
  • torch_signbit
  • torch_sparse_sampled_addmm
  • torch_sub
  • torch_subtract
  • torch_take_along_dim
  • torch_tensor_from_buffer
  • torch_unsafe_chunk
  • torch_unsafe_split
  • torch_vander
  • torch_vdot
  • torch_view_as_complex
  • torch_view_as_real
  • torch_vstack
  • unset_autocast
  • with_autocast
  • with_detect_anomaly
  • with_device
  • with_torch_manual_seed
  • yield

removed Removed (9)

  • torch_eig
  • torch_fft
  • torch_ifft
  • torch_irfft
  • torch_lstsq
  • torch_matrix_rank
  • torch_rfft
  • torch_solve
  • torch_symeig

Licence and people

Method, licences

Part0.0.30.17.0
changedMaintainerDaniel FalbelTomasz Kalinowski
addedPerson named–Christophe Regouby
addedPerson named–Hamada S. Badr
addedPerson named–Krzysztof Joachimiak
addedPerson named–Maximilian Pichler
addedPerson named–Sebastian Fischer
addedPerson named–Tomasz Kalinowski
changedDaniel Falbelaut, cre, cphaut, cph
changedJavier Luraschiaut, cphaut
Side by side
Measure0.0.30.17.0Change
LicenceMIT + file LICENSEMIT + file LICENSE
MaintainerDaniel FalbelTomasz Kalinowskichanged
People namedDaniel Falbel, Javier Luraschi, Dmitriy Selivanov, Athos Damiani, RStudioTomasz Kalinowski, Daniel Falbel, Javier Luraschi, Dmitriy Selivanov, Athos Damiani, Christophe Regouby, Krzysztof Joachimiak, Hamada S. Badr, Sebastian Fischer, Maximilian Pichler, RStudiochanged
Measure0.0.30.17.0Change
Lines of R code51,179102,571+51,392 up
Lines of compiled code43,63180,023+36,392 up
Files6881,037+349 up
Public exports494755+261 up
Internal functions2,7116,280+3,569 up
Highest complexity91155+64 up
Download size2.5 MB1.9 MB−620 KB down
Source, unpacked12.2 MB23.1 MB+10.9 MB up

1 other figure of the code is the same in both.

Side by side
Measure0.0.30.17.0Change
Lines of R code51,179102,571+51,392
Lines of compiled code43,63180,023+36,392
Files6881,037+349
Public exports494755+261
Internal functions2,7116,280+3,569
Mean complexity1.11.1
Highest complexity91155+64
Download size2.5 MB1.9 MB−620 KB
Source, unpacked12.2 MB23.1 MB+10.9 MB
Analysed withrpkg-analyzer 0.5.2rpkg-analyzer 0.5.2
Measure0.0.30.17.0Change
Lines run by testsNot measured0.3%
Test cases246735+489 up
Lines of tests3,41911,858+8,439 up
Test lines per line of code0.070.12+0.05 up
Exports named in tests26.3%46.0%+19.7 points up

Test coverage is measured on each CRAN package's newest release since July 2026.

Side by side
Measure0.0.30.17.0Change
Lines run by testsNot measured0.3%
Test cases246735+489
Lines of tests3,41911,858+8,439
Test lines per line of code0.070.12+0.05
Exports named in tests26.3%46.0%+19.7 points
Measure0.0.30.17.0Change
Exports documented99.8%99.9%+0.1 points up
Return value documented2.4%6.2%+3.8 points up
Lines of documentation19,06430,397+11,333 up
Lines of guides8,8783,872−5,006 down

1 other figure of the documentation is the same in both.

Side by side
Measure0.0.30.17.0Change
Exports documented99.8%99.9%+0.1 points
Arguments documented100.0%100.0%
Return value documented2.4%6.2%+3.8 points
Lines of documentation19,06430,397+11,333
Lines of guides8,8783,872−5,006

Files changed

Files changed
853over 29 releases
Lines added
802,781
Lines deleted
436,247
Most changed files (12 of 853)
FileLines addedLines deleted
tools/torchgen/inst/declaration/Declarations-2.8.0.yaml213,8510
tools/torchgen/inst/declaration/Declarations-1.7.0.yaml111,6160
src/RcppExports.cpp60,68333,645
tools/torchgen/inst/declaration/Declarations-1.12.1.yaml52,13932,081
tools/torchgen/inst/declaration/Declarations-1.9.0.yaml38,46230,660
tools/torchgen/inst/declaration/Declarations-1.5.0.yaml063,549
tools/torchgen/inst/declaration/Declarations-1.4.0.yaml060,108
tools/torchgen/inst/declaration/Declarations-1.3.0.yaml059,165
tools/torchgen/inst/declaration/Declarations-1.13.1.yaml44,8758,647
tools/torchgen/inst/declaration/Declarations-1.8.0.yaml29,61119,357
R/gen-namespace.R34,36111,441
src/gen-namespace.cpp27,45615,369

Each release is counted against the release before it, so a line changed twice counts twice.

Release notes

From the NEWS file of 0.17.0. Only the section for the newest release on record is kept.

  • Updated to LibTorch 2.8.0 (#1419).
  • torch_triu_indices() and torch_tril_indices() now return 1-based indexes (#1382).
  • $indices() now returns 1-based indexes (#1382).
  • Added torch_sitrep() for torch installation situation report (#1415, @cregouby).
  • Added %*% operator for torch tensors (#1379, @caio-hamamura).
  • Added head() and tail() methods for torch tensors (#1388).
  • Exported torch_scaled_dot_product_attention() (#1404).
  • Exported torch_sparse_sampled_addmm() (#1427).
  • Exported torch_ldexp() (#1407).
  • Integrated cudatoolkit R package for CUDA library loading (#1422).
  • Improved as_array() support for sparse tensors (#1387).
  • Fixed sparse CSR tensor printing, to_sparse() crash, and as_array() detection (#1431).
  • Fixed dataset()/nn_module() crash when initialize parameter is named d (#1424, @Chandraveer-Singh).
  • Fixed bfloat16 not being recognized as a floating point dtype (#1408).
  • Fixed handling of R longjump exceptions in autograd and tracing callbacks (#1406).
  • Fixed redundant lantern.h include causing macro redefinition warning (#1430).
  • Excluded glibc libraries from dependency bundling (#1397, @troyhernandez).
  • Replaced non-API entry point Rf_findVarInFrame with R_getVarEx for R 4.6 compatibility (#1421).
  • Added French translations for 5 missing messages (#1385, @cregouby).

Unchanged and not compared

Licence is the same in both releases. No advisory is known for either release.

Checks, the install tree, dependents and the repository describe the package as it is now, so they are on the package page.