Upgrading torchvision 0.3 to 0.4
Unknown0.3.0 → 0.4.2
Fetching 4 releases from GitHub…0 / 4
UnknownUpgrade verdict
DepDelta is still reading the notes for this range, so breaking changes can't be ruled out yet.
In plain terms
- DepDelta is still reading the release notes in this range, so it can't rule out breaking changes yet.
- This page updates as the notes come in.
Must act Still reading release notes
DepDelta is still reading the release notes in this range. This updates as they come in.
Installs 4 packages in all (was 32).
No longer installs cuda-bindings, cuda-pathfinder, cuda-toolkit, filelock, fsspec, jinja2, markupsafe, mpmath, networkx, nvidia-cublas, nvidia-cuda-cccl, nvidia-cuda-cupti, nvidia-cuda-nvrtc, nvidia-cuda-runtime, nvidia-cudnn-cu13 and 13 more.
Moves torch 2.15.0 → 1.3.1. Your build may pick other versions.
API changes
Generated by DepDelta from published Python source. Not mixed with prose notes.
- added:
export io - added:
function get_video_backend() - added:
function set_video_backend(backend)
- added:
export HMDB51 - added:
export Kinetics400 - added:
export QMNIST - added:
export UCF101 - added:
export USPS
- changed:
constructor Cityscapes(root, split=…, mode=…, target_type=…, transform=…, target_transform=…) → constructor Cityscapes(root, split=…, mode=…, target_type=…, transform=…, target_transform=…, transforms=…)
- added:
class HMDB51(VisionDataset) - added:
constructor HMDB51(root, annotation_path, frames_per_clip, step_between_clips=…, frame_rate=…, fold=…, train=…, transform=…, _precomputed_metadata=…, num_workers=…, _video_width=…, _video_height=…, _video_min_dimension=…, _audio_samples=…) - added:
field HMDB51.data_url - added:
field HMDB51.splits - added:
property HMDB51.metadata
- removed:
function download_and_extract_tar(url, download_root, extract_root=…, filename=…, md5=…, **kwargs) - removed:
function extract_tar(src, dest=…, gzip=…, delete=…) - removed:
property ImageNet.valid_splits
- added:
class Kinetics400(VisionDataset) - added:
constructor Kinetics400(root, frames_per_clip, step_between_clips=…, frame_rate=…, extensions=…, transform=…, _precomputed_metadata=…, num_workers=…, _video_width=…, _video_height=…, _video_min_dimension=…, _audio_samples=…) - added:
property Kinetics400.metadata
- removed:
static method MNIST.extract_gzip(gzip_path, remove_finished=…) - added:
class QMNIST(MNIST) - added:
constructor QMNIST(root, what=…, compat=…, train=…, **kwargs) - added:
field QMNIST.classes - added:
field QMNIST.subsets - added:
field QMNIST.urls - added:
function open_maybe_compressed_file(path) - added:
function read_sn3_pascalvincent_tensor(path, strict=…) - added:
method QMNIST.download() - added:
method QMNIST.extra_repr()
- added:
export DistributedSampler - added:
export RandomClipSampler - added:
export UniformClipSampler
- added:
class DistributedSampler(Sampler) - added:
class RandomClipSampler(torch.utils.data.Sampler) - added:
class UniformClipSampler(torch.utils.data.Sampler) - added:
constructor DistributedSampler(dataset, num_replicas=…, rank=…, shuffle=…, group_size=…) - added:
constructor RandomClipSampler(video_clips, max_clips_per_video) - added:
constructor UniformClipSampler(video_clips, num_clips_per_video) - added:
method DistributedSampler.set_epoch(epoch)
- changed:
class STL10(CIFAR10) → class STL10(VisionDataset) - changed:
constructor STL10(root, split=…, transform=…, target_transform=…, download=…) → constructor STL10(root, split=…, folds=…, transform=…, target_transform=…, download=…) - added:
field STL10.folds_list_file - added:
method STL10.download()
- added:
class UCF101(VisionDataset) - added:
constructor UCF101(root, annotation_path, frames_per_clip, step_between_clips=…, frame_rate=…, fold=…, train=…, transform=…, _precomputed_metadata=…, num_workers=…, _video_width=…, _video_height=…, _video_min_dimension=…, _audio_samples=…) - added:
property UCF101.metadata
- added:
class USPS(VisionDataset) - added:
constructor USPS(root, train=…, transform=…, target_transform=…, download=…) - added:
field USPS.split_list
- added:
function download_and_extract_archive(url, download_root, extract_root=…, filename=…, md5=…, remove_finished=…) - added:
function extract_archive(from_path, to_path=…, remove_finished=…) - added:
function iterable_to_str(iterable) - added:
function verify_str_arg(value, arg=…, valid_values=…, custom_msg=…)
- added:
class VideoClips(object) - added:
constructor VideoClips(video_paths, clip_length_in_frames=…, frames_between_clips=…, frame_rate=…, _precomputed_metadata=…, num_workers=…, _video_width=…, _video_height=…, _video_min_dimension=…, _audio_samples=…) - added:
function pts_convert(pts, timebase_from, timebase_to, round_func=…) - added:
function unfold(tensor, size, step, dilation=…) - added:
method VideoClips.compute_clips(num_frames, step, frame_rate=…) - added:
method VideoClips.get_clip(idx) - added:
method VideoClips.get_clip_location(idx) - added:
method VideoClips.num_clips() - added:
method VideoClips.num_videos() - added:
method VideoClips.subset(indices) - added:
property VideoClips.metadata - added:
static method VideoClips.compute_clips_for_video(video_pts, num_frames, step, fps, frame_rate)
- added:
export _HAS_VIDEO_OPT - added:
export _probe_video_from_file - added:
export _probe_video_from_memory - added:
export _read_video_from_file - added:
export _read_video_from_memory - added:
export _read_video_timestamps_from_file - added:
export _read_video_timestamps_from_memory - added:
export read_video - added:
export read_video_timestamps - added:
export write_video
- added:
function read_video(filename, start_pts=…, end_pts=…, pts_unit=…) - added:
function read_video_timestamps(filename, pts_unit=…) - added:
function write_video(filename, video_array, fps, video_codec=…, options=…) - added:
variable av - added:
variable description - added:
variable lib_dir - added:
variable path
- added:
export video
- changed:
constructor DenseNet(growth_rate=…, block_config=…, num_init_features=…, bn_size=…, drop_rate=…, num_classes=…) → constructor DenseNet(growth_rate=…, block_config=…, num_init_features=…, bn_size=…, drop_rate=…, num_classes=…, memory_efficient=…)
- added:
class MNASNet(torch.nn.Module) - added:
constructor MNASNet(alpha, num_classes=…, dropout=…) - added:
function mnasnet0_5(pretrained=…, progress=…, **kwargs) - added:
function mnasnet0_75(pretrained=…, progress=…, **kwargs) - added:
function mnasnet1_0(pretrained=…, progress=…, **kwargs) - added:
function mnasnet1_3(pretrained=…, progress=…, **kwargs) - added:
method MNASNet.forward(x)
- changed:
constructor MobileNetV2(num_classes=…, width_mult=…) → constructor MobileNetV2(num_classes=…, width_mult=…, inverted_residual_setting=…, round_nearest=…)
- added:
function wide_resnet101_2(pretrained=…, progress=…, **kwargs) - added:
function wide_resnet50_2(pretrained=…, progress=…, **kwargs)
- added:
function mc3_18(pretrained=…, progress=…, **kwargs) - added:
function r2plus1d_18(pretrained=…, progress=…, **kwargs) - added:
function r3d_18(pretrained=…, progress=…, **kwargs)
- added:
function erase(img, i, j, h, w, v, inplace=…)
- added:
class RandomErasing(object) - added:
constructor RandomErasing(p=…, scale=…, ratio=…, value=…, inplace=…) - added:
static method RandomErasing.get_params(img, scale, ratio, value=…)