dist_train_arm.sh
for ARM device and update NPU results. (#1218)MMClsWandbHook
stuck. (#1242)device_ids
in tools/test.py
. (#1215)pytorch2torchscript.md
. (#1173)miscellaneous.md
. (#1137)ClassBalancedDataset
. (#901)get_sinusoid_encoding
from third-party implementation. (#965)RepVGG
. (#985)train_step
instead of forward
in PreciseBNHook (#964)forward_dummy
to calculate FLOPS. (#953)torch.meshgrid
. (#860)--out-dir
option of analyze_results.py. (#898)--local_rank
.SwinTransformer
. (#749)CustomDataset
class to help you build dataset of yourself!CustomDataset
. (#738)dist_train
& dist_test
tools support distributed training on multiple machines. (#734)--a-b
instead of --a_b
in arguments. (#754)get_cat_ids
and get_gt_labels
to KFoldDataset. (#721)worker_init_fn
. (#733)evaluate
function for ConcatDataset. (#650)one_hot
to implement convert_to_one_hot
. (#696)--gpu-id
instead of --gpu-ids
in non-distributed multi-gpu training/testing. (#688)dist.barrier
to fix a bug in directory checking. (#666)features_only
option in TIMMBackbone
. (#668)metric_options
is not specified in multi-label evaluation. (#647)pytorch-grad-cam>=1.3.7
. (#656)cal_train_time
in analyze_logs.py
. (#662)NumClassCheckHook
and unit tests. (#559)analysis_log.py
. (#529)import_modules_from_string
. (#544)reset_classifier
to remove head of timm backbones. (#534)Resize
transform and add Pad
transform. (#506)ClassBalancedDataset
. (#555)Resize
. (#547)analyze_result.py
. (#518)analyze_logs.py
and prevent empty curve. (#510)visualization.md
and add example images. (#513)DistSamplerSeedHook
if use IterBasedRunner
. (#501)EvalHook
to "LOW" to avoid a bug when using IterBasedRunner
. (#488)get_root_logger
in apis/train.py
. (#486)--out-items
in tools/test.py
. (#437)--options
to --cfg-options
in some tools. (#425)test.py
when metric returns np.ndarray
. (#441)publish_model
bug if no parent of out_file
. (#463)packaging
. (#459)getting_started.md
and install.md
. And rewrite finetune.md
. (#466)CITATION.cff
. (#428)hparams
argument in AutoAugment
and RandAugment
to provide hyperparameters for sub-policies.SELayer
.hparams
argument in AutoAugment
and RandAugment
and some other improvement. (#398)SELayer
to support custom squeeze channels. (#417)post_process
function to handle pred result processing. (#390)digit_version
function. (#402)presistent_works
option if available, to accelerate training. (#349)thrs
in metrics. (#341)patch_cfg
argument bug in SwinTransformer. (#368)init_weights
call in ViT init function. (#373)_base_
link in a resnet config. (#361)CONTRIBUTING.md
and all tools tutorials. (#320)PatchEmbed
and HybridEmbed
as independent components. (#330)Augments
to support more functions. (#278)LabelSmoothLoss
to support multiple calculation formulas. (#285)num_imgs
can not be evenly divided by num_gpus
. (#299)magnitude_std
bug in RandAugment
. (#309)samples_per_gpu
is 1. (#311)data_pipeline.md
and new_modules.md
. (#265)RandomResizedCrop
and CenterCrop
. (#268)base_head
. (#274)data.test
in MNIST configs. (#264)new_dataset.md
and add Chinese translation of finture.md
, new_dataset.md
.dim
argument for GlobalAveragePooling
. (#236)RandAugment
magnitude. (#240)new_dataset.md
and add Chinese translation of finture.md
, new_dataset.md
. (#243)magnitude_range
in RandAugment
. (#249)analyze_results.py
. (#237)MANIFEST.in
. (#250 & #255)tools/deployment/test.py
as a ONNX runtime test tool.README.md
and some Chinese tutorials.simplify
option in pytorch2onnx.py
. (#200)tools/deployment/test.py
as a ONNX runtime test tool. (#212)device
option to support training on CPU. (#219)README.md
and some Chinese tutorials. (#221)metafile.yml
in configs to support interaction with paper with code(PWC) and MMCLI. (#225)LabelSmoothLoss
so that label smoothing and mixup could be enabled at the same time. (#203)cal_acc
option in ClsHead
. (#206)CLASSES
in checkpoint to avoid unexpected key error. (#207)CONTRIBUTING.md
to align with that in MMCV. (#210)pytorch2onnx.md
tutorial. (#229)setup.py
to support MMCLI. (#232)Rotate
pipeline for data augmentation. (#167)Invert
pipeline for data augmentation. (#168)Color
pipeline for data augmentation. (#171)Solarize
and Posterize
pipeline for data augmentation. (#172)AutoAugmentation
, AutoContrast
, Equalize
, Contrast
, Brightness
and Sharpness
pipelines for data augmentation. (#179)Shear
pipeline for data augmentation. (#163)Translate
pipeline for data augmentation. (#165)tools/onnx2tensorrt.py
as a tool to create TensorRT engine from ONNX, run inference and verify outputs in Python. (#153)--eval-options
in tools/test.py
to support eval options override, matching the behavior of other open-mmlab projects. (#158)mmcls.apis.test
and tools/test.py
, matching the behavior of other open-mmlab projects. (#162)RandomCrop
and RandomResizedCrop
. (#151)meta_keys
in Collect
. (#149 & #152)--options
. (#91 & #96)build_runner
to make runners more flexible. (#54)BaseDataset
. (#72)CLASSES
override during BaseDateset
initialization. (#85)CLASSES
argument to dataset wrappers. (#66)Accuracy
. (#104)gpu_ids
in distributed training. (#107)Resize
. (#21)Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》