Full Code of salesforce/LAVIS for AI

gitextract__ei5npya/

├── .github/
│   └── workflows/
│       └── docs.yaml
├── .gitignore
├── .pre-commit-config.yaml
├── CODEOWNERS
├── CODE_OF_CONDUCT.md
├── LICENSE.txt
├── MANIFEST.in
├── README.md
├── SECURITY.md
├── app/
│   ├── __init__.py
│   ├── calculate_coco_features.py
│   ├── caption.py
│   ├── classification.py
│   ├── dataset_browser.py
│   ├── image_text_match.py
│   ├── main.py
│   ├── multimodal_search.py
│   ├── multipage.py
│   ├── text_localization.py
│   ├── utils.py
│   └── vqa.py
├── dataset_card/
│   ├── avsd_dialogue.md
│   ├── coco_caption.md
│   ├── coco_retrieval.md
│   ├── conceptual_captions.md
│   ├── didemo_retrieval.md
│   ├── flickr_retrieval.md
│   ├── gqa.md
│   ├── msrvtt_qa.md
│   ├── msrvtt_retrieval.md
│   ├── msvd_qa.md
│   ├── nlvr2.md
│   ├── nocaps.md
│   ├── sbu_caption.md
│   ├── snli_visual_entailment.md
│   └── vqav2.md
├── docs/
│   ├── Makefile
│   ├── benchmark.rst
│   ├── build_docs.sh
│   ├── conf.py
│   ├── getting_started.rst
│   ├── index.rst
│   ├── intro.rst
│   ├── make.bat
│   ├── requirements.txt
│   ├── tutorial.configs.rst
│   ├── tutorial.datasets.rst
│   ├── tutorial.evaluation.rst
│   ├── tutorial.models.rst
│   ├── tutorial.processors.rst
│   ├── tutorial.rst
│   ├── tutorial.tasks.rst
│   └── tutorial.training-example.rst
├── evaluate.py
├── examples/
│   ├── albef_feature_extraction.ipynb
│   ├── albef_vqa.ipynb
│   ├── albef_zero_shot_classification.ipynb
│   ├── blip2_feature_extraction.ipynb
│   ├── blip2_image_text_matching.ipynb
│   ├── blip2_instructed_generation.ipynb
│   ├── blip_feature_extraction.ipynb
│   ├── blip_image_captioning.ipynb
│   ├── blip_image_text_matching.ipynb
│   ├── blip_text_localization.ipynb
│   ├── blip_vqa.ipynb
│   ├── blip_zero_shot_classification.ipynb
│   ├── clip_feature_extraction.ipynb
│   └── clip_zero_shot_classification.ipynb
├── lavis/
│   ├── __init__.py
│   ├── common/
│   │   ├── annotator/
│   │   │   ├── canny/
│   │   │   │   └── __init__.py
│   │   │   ├── ckpts/
│   │   │   │   └── download.sh
│   │   │   ├── hed/
│   │   │   │   └── __init__.py
│   │   │   ├── midas/
│   │   │   │   ├── __init__.py
│   │   │   │   ├── api.py
│   │   │   │   ├── midas/
│   │   │   │   │   ├── __init__.py
│   │   │   │   │   ├── base_model.py
│   │   │   │   │   ├── blocks.py
│   │   │   │   │   ├── dpt_depth.py
│   │   │   │   │   ├── midas_net.py
│   │   │   │   │   ├── midas_net_custom.py
│   │   │   │   │   ├── transforms.py
│   │   │   │   │   └── vit.py
│   │   │   │   └── utils.py
│   │   │   ├── mlsd/
│   │   │   │   ├── __init__.py
│   │   │   │   ├── models/
│   │   │   │   │   ├── mbv2_mlsd_large.py
│   │   │   │   │   └── mbv2_mlsd_tiny.py
│   │   │   │   └── utils.py
│   │   │   ├── openpose/
│   │   │   │   ├── __init__.py
│   │   │   │   ├── body.py
│   │   │   │   ├── hand.py
│   │   │   │   ├── model.py
│   │   │   │   └── util.py
│   │   │   ├── uniformer/
│   │   │   │   ├── __init__.py
│   │   │   │   ├── configs/
│   │   │   │   │   └── _base_/
│   │   │   │   │       ├── datasets/
│   │   │   │   │       │   ├── ade20k.py
│   │   │   │   │       │   ├── chase_db1.py
│   │   │   │   │       │   ├── cityscapes.py
│   │   │   │   │       │   ├── cityscapes_769x769.py
│   │   │   │   │       │   ├── drive.py
│   │   │   │   │       │   ├── hrf.py
│   │   │   │   │       │   ├── pascal_context.py
│   │   │   │   │       │   ├── pascal_context_59.py
│   │   │   │   │       │   ├── pascal_voc12.py
│   │   │   │   │       │   ├── pascal_voc12_aug.py
│   │   │   │   │       │   └── stare.py
│   │   │   │   │       ├── default_runtime.py
│   │   │   │   │       ├── models/
│   │   │   │   │       │   ├── ann_r50-d8.py
│   │   │   │   │       │   ├── apcnet_r50-d8.py
│   │   │   │   │       │   ├── ccnet_r50-d8.py
│   │   │   │   │       │   ├── cgnet.py
│   │   │   │   │       │   ├── danet_r50-d8.py
│   │   │   │   │       │   ├── deeplabv3_r50-d8.py
│   │   │   │   │       │   ├── deeplabv3_unet_s5-d16.py
│   │   │   │   │       │   ├── deeplabv3plus_r50-d8.py
│   │   │   │   │       │   ├── dmnet_r50-d8.py
│   │   │   │   │       │   ├── dnl_r50-d8.py
│   │   │   │   │       │   ├── emanet_r50-d8.py
│   │   │   │   │       │   ├── encnet_r50-d8.py
│   │   │   │   │       │   ├── fast_scnn.py
│   │   │   │   │       │   ├── fcn_hr18.py
│   │   │   │   │       │   ├── fcn_r50-d8.py
│   │   │   │   │       │   ├── fcn_unet_s5-d16.py
│   │   │   │   │       │   ├── fpn_r50.py
│   │   │   │   │       │   ├── fpn_uniformer.py
│   │   │   │   │       │   ├── gcnet_r50-d8.py
│   │   │   │   │       │   ├── lraspp_m-v3-d8.py
│   │   │   │   │       │   ├── nonlocal_r50-d8.py
│   │   │   │   │       │   ├── ocrnet_hr18.py
│   │   │   │   │       │   ├── ocrnet_r50-d8.py
│   │   │   │   │       │   ├── pointrend_r50.py
│   │   │   │   │       │   ├── psanet_r50-d8.py
│   │   │   │   │       │   ├── pspnet_r50-d8.py
│   │   │   │   │       │   ├── pspnet_unet_s5-d16.py
│   │   │   │   │       │   ├── upernet_r50.py
│   │   │   │   │       │   └── upernet_uniformer.py
│   │   │   │   │       └── schedules/
│   │   │   │   │           ├── schedule_160k.py
│   │   │   │   │           ├── schedule_20k.py
│   │   │   │   │           ├── schedule_40k.py
│   │   │   │   │           └── schedule_80k.py
│   │   │   │   ├── exp/
│   │   │   │   │   └── upernet_global_small/
│   │   │   │   │       ├── config.py
│   │   │   │   │       ├── run.sh
│   │   │   │   │       ├── test.sh
│   │   │   │   │       ├── test_config_g.py
│   │   │   │   │       ├── test_config_h32.py
│   │   │   │   │       └── test_config_w32.py
│   │   │   │   ├── mmcv/
│   │   │   │   │   ├── __init__.py
│   │   │   │   │   ├── arraymisc/
│   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   └── quantization.py
│   │   │   │   │   ├── cnn/
│   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   ├── alexnet.py
│   │   │   │   │   │   ├── bricks/
│   │   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   │   ├── activation.py
│   │   │   │   │   │   │   ├── context_block.py
│   │   │   │   │   │   │   ├── conv.py
│   │   │   │   │   │   │   ├── conv2d_adaptive_padding.py
│   │   │   │   │   │   │   ├── conv_module.py
│   │   │   │   │   │   │   ├── conv_ws.py
│   │   │   │   │   │   │   ├── depthwise_separable_conv_module.py
│   │   │   │   │   │   │   ├── drop.py
│   │   │   │   │   │   │   ├── generalized_attention.py
│   │   │   │   │   │   │   ├── hsigmoid.py
│   │   │   │   │   │   │   ├── hswish.py
│   │   │   │   │   │   │   ├── non_local.py
│   │   │   │   │   │   │   ├── norm.py
│   │   │   │   │   │   │   ├── padding.py
│   │   │   │   │   │   │   ├── plugin.py
│   │   │   │   │   │   │   ├── registry.py
│   │   │   │   │   │   │   ├── scale.py
│   │   │   │   │   │   │   ├── swish.py
│   │   │   │   │   │   │   ├── transformer.py
│   │   │   │   │   │   │   ├── upsample.py
│   │   │   │   │   │   │   └── wrappers.py
│   │   │   │   │   │   ├── builder.py
│   │   │   │   │   │   ├── resnet.py
│   │   │   │   │   │   ├── utils/
│   │   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   │   ├── flops_counter.py
│   │   │   │   │   │   │   ├── fuse_conv_bn.py
│   │   │   │   │   │   │   ├── sync_bn.py
│   │   │   │   │   │   │   └── weight_init.py
│   │   │   │   │   │   └── vgg.py
│   │   │   │   │   ├── engine/
│   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   └── test.py
│   │   │   │   │   ├── fileio/
│   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   ├── file_client.py
│   │   │   │   │   │   ├── handlers/
│   │   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   │   ├── base.py
│   │   │   │   │   │   │   ├── json_handler.py
│   │   │   │   │   │   │   ├── pickle_handler.py
│   │   │   │   │   │   │   └── yaml_handler.py
│   │   │   │   │   │   ├── io.py
│   │   │   │   │   │   └── parse.py
│   │   │   │   │   ├── image/
│   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   ├── colorspace.py
│   │   │   │   │   │   ├── geometric.py
│   │   │   │   │   │   ├── io.py
│   │   │   │   │   │   ├── misc.py
│   │   │   │   │   │   └── photometric.py
│   │   │   │   │   ├── model_zoo/
│   │   │   │   │   │   ├── deprecated.json
│   │   │   │   │   │   ├── mmcls.json
│   │   │   │   │   │   └── open_mmlab.json
│   │   │   │   │   ├── ops/
│   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   ├── assign_score_withk.py
│   │   │   │   │   │   ├── ball_query.py
│   │   │   │   │   │   ├── bbox.py
│   │   │   │   │   │   ├── border_align.py
│   │   │   │   │   │   ├── box_iou_rotated.py
│   │   │   │   │   │   ├── carafe.py
│   │   │   │   │   │   ├── cc_attention.py
│   │   │   │   │   │   ├── contour_expand.py
│   │   │   │   │   │   ├── corner_pool.py
│   │   │   │   │   │   ├── correlation.py
│   │   │   │   │   │   ├── deform_conv.py
│   │   │   │   │   │   ├── deform_roi_pool.py
│   │   │   │   │   │   ├── deprecated_wrappers.py
│   │   │   │   │   │   ├── focal_loss.py
│   │   │   │   │   │   ├── furthest_point_sample.py
│   │   │   │   │   │   ├── fused_bias_leakyrelu.py
│   │   │   │   │   │   ├── gather_points.py
│   │   │   │   │   │   ├── group_points.py
│   │   │   │   │   │   ├── info.py
│   │   │   │   │   │   ├── iou3d.py
│   │   │   │   │   │   ├── knn.py
│   │   │   │   │   │   ├── masked_conv.py
│   │   │   │   │   │   ├── merge_cells.py
│   │   │   │   │   │   ├── modulated_deform_conv.py
│   │   │   │   │   │   ├── multi_scale_deform_attn.py
│   │   │   │   │   │   ├── nms.py
│   │   │   │   │   │   ├── pixel_group.py
│   │   │   │   │   │   ├── point_sample.py
│   │   │   │   │   │   ├── points_in_boxes.py
│   │   │   │   │   │   ├── points_sampler.py
│   │   │   │   │   │   ├── psa_mask.py
│   │   │   │   │   │   ├── roi_align.py
│   │   │   │   │   │   ├── roi_align_rotated.py
│   │   │   │   │   │   ├── roi_pool.py
│   │   │   │   │   │   ├── roiaware_pool3d.py
│   │   │   │   │   │   ├── roipoint_pool3d.py
│   │   │   │   │   │   ├── saconv.py
│   │   │   │   │   │   ├── scatter_points.py
│   │   │   │   │   │   ├── sync_bn.py
│   │   │   │   │   │   ├── three_interpolate.py
│   │   │   │   │   │   ├── three_nn.py
│   │   │   │   │   │   ├── tin_shift.py
│   │   │   │   │   │   ├── upfirdn2d.py
│   │   │   │   │   │   └── voxelize.py
│   │   │   │   │   ├── parallel/
│   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   ├── _functions.py
│   │   │   │   │   │   ├── collate.py
│   │   │   │   │   │   ├── data_container.py
│   │   │   │   │   │   ├── data_parallel.py
│   │   │   │   │   │   ├── distributed.py
│   │   │   │   │   │   ├── distributed_deprecated.py
│   │   │   │   │   │   ├── registry.py
│   │   │   │   │   │   ├── scatter_gather.py
│   │   │   │   │   │   └── utils.py
│   │   │   │   │   ├── runner/
│   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   ├── base_module.py
│   │   │   │   │   │   ├── base_runner.py
│   │   │   │   │   │   ├── builder.py
│   │   │   │   │   │   ├── checkpoint.py
│   │   │   │   │   │   ├── default_constructor.py
│   │   │   │   │   │   ├── dist_utils.py
│   │   │   │   │   │   ├── epoch_based_runner.py
│   │   │   │   │   │   ├── fp16_utils.py
│   │   │   │   │   │   ├── hooks/
│   │   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   │   ├── checkpoint.py
│   │   │   │   │   │   │   ├── closure.py
│   │   │   │   │   │   │   ├── ema.py
│   │   │   │   │   │   │   ├── evaluation.py
│   │   │   │   │   │   │   ├── hook.py
│   │   │   │   │   │   │   ├── iter_timer.py
│   │   │   │   │   │   │   ├── logger/
│   │   │   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   │   │   ├── base.py
│   │   │   │   │   │   │   │   ├── dvclive.py
│   │   │   │   │   │   │   │   ├── mlflow.py
│   │   │   │   │   │   │   │   ├── neptune.py
│   │   │   │   │   │   │   │   ├── pavi.py
│   │   │   │   │   │   │   │   ├── tensorboard.py
│   │   │   │   │   │   │   │   ├── text.py
│   │   │   │   │   │   │   │   └── wandb.py
│   │   │   │   │   │   │   ├── lr_updater.py
│   │   │   │   │   │   │   ├── memory.py
│   │   │   │   │   │   │   ├── momentum_updater.py
│   │   │   │   │   │   │   ├── optimizer.py
│   │   │   │   │   │   │   ├── profiler.py
│   │   │   │   │   │   │   ├── sampler_seed.py
│   │   │   │   │   │   │   └── sync_buffer.py
│   │   │   │   │   │   ├── iter_based_runner.py
│   │   │   │   │   │   ├── log_buffer.py
│   │   │   │   │   │   ├── optimizer/
│   │   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   │   ├── builder.py
│   │   │   │   │   │   │   └── default_constructor.py
│   │   │   │   │   │   ├── priority.py
│   │   │   │   │   │   └── utils.py
│   │   │   │   │   ├── utils/
│   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   ├── config.py
│   │   │   │   │   │   ├── env.py
│   │   │   │   │   │   ├── ext_loader.py
│   │   │   │   │   │   ├── logging.py
│   │   │   │   │   │   ├── misc.py
│   │   │   │   │   │   ├── parrots_jit.py
│   │   │   │   │   │   ├── parrots_wrapper.py
│   │   │   │   │   │   ├── path.py
│   │   │   │   │   │   ├── progressbar.py
│   │   │   │   │   │   ├── registry.py
│   │   │   │   │   │   ├── testing.py
│   │   │   │   │   │   ├── timer.py
│   │   │   │   │   │   ├── trace.py
│   │   │   │   │   │   └── version_utils.py
│   │   │   │   │   ├── version.py
│   │   │   │   │   ├── video/
│   │   │   │   │   │   ├── __init__.py
│   │   │   │   │   │   ├── io.py
│   │   │   │   │   │   ├── optflow.py
│   │   │   │   │   │   └── processing.py
│   │   │   │   │   └── visualization/
│   │   │   │   │       ├── __init__.py
│   │   │   │   │       ├── color.py
│   │   │   │   │       ├── image.py
│   │   │   │   │       └── optflow.py
│   │   │   │   ├── mmcv_custom/
│   │   │   │   │   ├── __init__.py
│   │   │   │   │   └── checkpoint.py
│   │   │   │   └── mmseg/
│   │   │   │       ├── apis/
│   │   │   │       │   ├── __init__.py
│   │   │   │       │   ├── inference.py
│   │   │   │       │   ├── test.py
│   │   │   │       │   └── train.py
│   │   │   │       ├── core/
│   │   │   │       │   ├── __init__.py
│   │   │   │       │   ├── evaluation/
│   │   │   │       │   │   ├── __init__.py
│   │   │   │       │   │   ├── class_names.py
│   │   │   │       │   │   ├── eval_hooks.py
│   │   │   │       │   │   └── metrics.py
│   │   │   │       │   ├── seg/
│   │   │   │       │   │   ├── __init__.py
│   │   │   │       │   │   ├── builder.py
│   │   │   │       │   │   └── sampler/
│   │   │   │       │   │       ├── __init__.py
│   │   │   │       │   │       ├── base_pixel_sampler.py
│   │   │   │       │   │       └── ohem_pixel_sampler.py
│   │   │   │       │   └── utils/
│   │   │   │       │       ├── __init__.py
│   │   │   │       │       └── misc.py
│   │   │   │       ├── datasets/
│   │   │   │       │   ├── __init__.py
│   │   │   │       │   ├── ade.py
│   │   │   │       │   ├── builder.py
│   │   │   │       │   ├── chase_db1.py
│   │   │   │       │   ├── cityscapes.py
│   │   │   │       │   ├── custom.py
│   │   │   │       │   ├── dataset_wrappers.py
│   │   │   │       │   ├── drive.py
│   │   │   │       │   ├── hrf.py
│   │   │   │       │   ├── pascal_context.py
│   │   │   │       │   ├── pipelines/
│   │   │   │       │   │   ├── __init__.py
│   │   │   │       │   │   ├── compose.py
│   │   │   │       │   │   ├── formating.py
│   │   │   │       │   │   ├── loading.py
│   │   │   │       │   │   ├── test_time_aug.py
│   │   │   │       │   │   └── transforms.py
│   │   │   │       │   ├── stare.py
│   │   │   │       │   └── voc.py
│   │   │   │       ├── models/
│   │   │   │       │   ├── __init__.py
│   │   │   │       │   ├── backbones/
│   │   │   │       │   │   ├── __init__.py
│   │   │   │       │   │   ├── cgnet.py
│   │   │   │       │   │   ├── fast_scnn.py
│   │   │   │       │   │   ├── hrnet.py
│   │   │   │       │   │   ├── mobilenet_v2.py
│   │   │   │       │   │   ├── mobilenet_v3.py
│   │   │   │       │   │   ├── resnest.py
│   │   │   │       │   │   ├── resnet.py
│   │   │   │       │   │   ├── resnext.py
│   │   │   │       │   │   ├── unet.py
│   │   │   │       │   │   ├── uniformer.py
│   │   │   │       │   │   └── vit.py
│   │   │   │       │   ├── builder.py
│   │   │   │       │   ├── decode_heads/
│   │   │   │       │   │   ├── __init__.py
│   │   │   │       │   │   ├── ann_head.py
│   │   │   │       │   │   ├── apc_head.py
│   │   │   │       │   │   ├── aspp_head.py
│   │   │   │       │   │   ├── cascade_decode_head.py
│   │   │   │       │   │   ├── cc_head.py
│   │   │   │       │   │   ├── da_head.py
│   │   │   │       │   │   ├── decode_head.py
│   │   │   │       │   │   ├── dm_head.py
│   │   │   │       │   │   ├── dnl_head.py
│   │   │   │       │   │   ├── ema_head.py
│   │   │   │       │   │   ├── enc_head.py
│   │   │   │       │   │   ├── fcn_head.py
│   │   │   │       │   │   ├── fpn_head.py
│   │   │   │       │   │   ├── gc_head.py
│   │   │   │       │   │   ├── lraspp_head.py
│   │   │   │       │   │   ├── nl_head.py
│   │   │   │       │   │   ├── ocr_head.py
│   │   │   │       │   │   ├── point_head.py
│   │   │   │       │   │   ├── psa_head.py
│   │   │   │       │   │   ├── psp_head.py
│   │   │   │       │   │   ├── sep_aspp_head.py
│   │   │   │       │   │   ├── sep_fcn_head.py
│   │   │   │       │   │   └── uper_head.py
│   │   │   │       │   ├── losses/
│   │   │   │       │   │   ├── __init__.py
│   │   │   │       │   │   ├── accuracy.py
│   │   │   │       │   │   ├── cross_entropy_loss.py
│   │   │   │       │   │   ├── dice_loss.py
│   │   │   │       │   │   ├── lovasz_loss.py
│   │   │   │       │   │   └── utils.py
│   │   │   │       │   ├── necks/
│   │   │   │       │   │   ├── __init__.py
│   │   │   │       │   │   ├── fpn.py
│   │   │   │       │   │   └── multilevel_neck.py
│   │   │   │       │   ├── segmentors/
│   │   │   │       │   │   ├── __init__.py
│   │   │   │       │   │   ├── base.py
│   │   │   │       │   │   ├── cascade_encoder_decoder.py
│   │   │   │       │   │   └── encoder_decoder.py
│   │   │   │       │   └── utils/
│   │   │   │       │       ├── __init__.py
│   │   │   │       │       ├── drop.py
│   │   │   │       │       ├── inverted_residual.py
│   │   │   │       │       ├── make_divisible.py
│   │   │   │       │       ├── res_layer.py
│   │   │   │       │       ├── se_layer.py
│   │   │   │       │       ├── self_attention_block.py
│   │   │   │       │       ├── up_conv_block.py
│   │   │   │       │       └── weight_init.py
│   │   │   │       ├── ops/
│   │   │   │       │   ├── __init__.py
│   │   │   │       │   ├── encoding.py
│   │   │   │       │   └── wrappers.py
│   │   │   │       └── utils/
│   │   │   │           ├── __init__.py
│   │   │   │           ├── collect_env.py
│   │   │   │           └── logger.py
│   │   │   └── util.py
│   │   ├── config.py
│   │   ├── dist_utils.py
│   │   ├── gradcam.py
│   │   ├── logger.py
│   │   ├── optims.py
│   │   ├── registry.py
│   │   ├── utils.py
│   │   └── vqa_tools/
│   │       ├── __init__.py
│   │       ├── vqa.py
│   │       └── vqa_eval.py
│   ├── configs/
│   │   ├── datasets/
│   │   │   ├── aokvqa/
│   │   │   │   ├── defaults.yaml
│   │   │   │   └── defaults_instruct.yaml
│   │   │   ├── audiocaps/
│   │   │   │   ├── defaults_mm_cap.yaml
│   │   │   │   ├── defaults_mm_cap_instruct.yaml
│   │   │   │   └── defaults_mm_qa.yaml
│   │   │   ├── audioset/
│   │   │   │   ├── defaults_mm_cap.yaml
│   │   │   │   └── defaults_mm_cap_instruct.yaml
│   │   │   ├── avsd/
│   │   │   │   ├── defaults_dial.yaml
│   │   │   │   └── defaults_mm_dial_instruct.yaml
│   │   │   ├── blip_diffusion_datasets/
│   │   │   │   └── defaults.yaml
│   │   │   ├── capfilt14m/
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   └── defaults_cap_instruct.yaml
│   │   │   ├── charade/
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   └── defaults_cap_instruct.yaml
│   │   │   ├── clotho/
│   │   │   │   ├── defaults_mm_cap.yaml
│   │   │   │   ├── defaults_mm_cap_instruct.yaml
│   │   │   │   └── defaults_mm_qa.yaml
│   │   │   ├── coco/
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   ├── defaults_cap_instruct.yaml
│   │   │   │   ├── defaults_ret.yaml
│   │   │   │   ├── defaults_vqa.yaml
│   │   │   │   ├── defaults_vqa_instruct.yaml
│   │   │   │   └── eval_vqa.yaml
│   │   │   ├── coin/
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   └── defaults_cap_instruct.yaml
│   │   │   ├── conceptual_caption/
│   │   │   │   ├── defaults_12m.yaml
│   │   │   │   ├── defaults_12m_instruct.yaml
│   │   │   │   ├── defaults_3m.yaml
│   │   │   │   └── defaults_3m_instruct.yaml
│   │   │   ├── didemo/
│   │   │   │   └── defaults_ret.yaml
│   │   │   ├── discriminatory_reasoning/
│   │   │   │   ├── defaults_mm_audio_video.yaml
│   │   │   │   ├── defaults_mm_image_pc.yaml
│   │   │   │   └── discriminatory_dataset/
│   │   │   │       ├── audiocaps_discrn.json
│   │   │   │       └── objaverse_discrn.json
│   │   │   ├── esc50/
│   │   │   │   └── defaults_mm_cls.yaml
│   │   │   ├── flickr30k/
│   │   │   │   ├── defaults.yaml
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   └── defaults_cap_instruct.yaml
│   │   │   ├── gqa/
│   │   │   │   ├── balanced_testdev.yaml
│   │   │   │   ├── balanced_testdev_instruct.yaml
│   │   │   │   ├── balanced_val.yaml
│   │   │   │   ├── balanced_val_instruct.yaml
│   │   │   │   ├── defaults.yaml
│   │   │   │   └── defaults_instruct.yaml
│   │   │   ├── iconqa/
│   │   │   │   ├── defaults.yaml
│   │   │   │   └── defaults_instruct.yaml
│   │   │   ├── imagenet/
│   │   │   │   └── defaults.yaml
│   │   │   ├── laion/
│   │   │   │   ├── defaults_2B_multi.yaml
│   │   │   │   ├── defaults_400M.yaml
│   │   │   │   └── defaults_400M_instruct.yaml
│   │   │   ├── llava150k/
│   │   │   │   └── defaults_dial.yaml
│   │   │   ├── modelnet40/
│   │   │   │   └── defaults_cls.yaml
│   │   │   ├── msrvtt/
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   ├── defaults_cap_instruct.yaml
│   │   │   │   ├── defaults_qa.yaml
│   │   │   │   ├── defaults_qa_instruct.yaml
│   │   │   │   └── defaults_ret.yaml
│   │   │   ├── msvd/
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   ├── defaults_cap_instruct.yaml
│   │   │   │   ├── defaults_qa.yaml
│   │   │   │   └── defaults_qa_instruct.yaml
│   │   │   ├── music_avqa/
│   │   │   │   ├── defaults_mm_qa.yaml
│   │   │   │   └── defaults_mm_qa_instruct.yaml
│   │   │   ├── nlvr/
│   │   │   │   └── defaults.yaml
│   │   │   ├── nocaps/
│   │   │   │   └── defaults.yaml
│   │   │   ├── objaverse/
│   │   │   │   ├── defaults_mm_cap.yaml
│   │   │   │   ├── defaults_mm_cap_instruct.yaml
│   │   │   │   └── defaults_mm_qa.yaml
│   │   │   ├── ocrvqa/
│   │   │   │   ├── defaults.yaml
│   │   │   │   └── defaults_instruct.yaml
│   │   │   ├── okvqa/
│   │   │   │   ├── defaults.yaml
│   │   │   │   └── defaults_instruct.yaml
│   │   │   ├── sbu_caption/
│   │   │   │   ├── defaults.yaml
│   │   │   │   └── defaults_instruct.yaml
│   │   │   ├── scienceqa/
│   │   │   │   ├── defaults.yaml
│   │   │   │   └── defaults_instruct.yaml
│   │   │   ├── shapenet/
│   │   │   │   ├── defaults_mm_cap.yaml
│   │   │   │   └── defaults_mm_cap_instruct.yaml
│   │   │   ├── snli_ve/
│   │   │   │   ├── defaults.yaml
│   │   │   │   └── defaults_instruct.yaml
│   │   │   ├── textcaps/
│   │   │   │   ├── defaults.yaml
│   │   │   │   └── defaults_instruct.yaml
│   │   │   ├── valor/
│   │   │   │   ├── defaults_mm_cap.yaml
│   │   │   │   └── defaults_mm_cap_instruct.yaml
│   │   │   ├── vatex/
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   └── defaults_cap_instruct.yaml
│   │   │   ├── vg/
│   │   │   │   ├── defaults_caption.yaml
│   │   │   │   ├── defaults_caption_instruct.yaml
│   │   │   │   ├── defaults_vqa.yaml
│   │   │   │   └── defaults_vqa_instruct.yaml
│   │   │   ├── violin/
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   ├── defaults_cap_instruct.yaml
│   │   │   │   ├── defaults_entail.yaml
│   │   │   │   └── defaults_entail_instruct.yaml
│   │   │   ├── visdial/
│   │   │   │   ├── defaults_dial.yaml
│   │   │   │   └── defaults_dial_instruct.yaml
│   │   │   ├── vizwiz/
│   │   │   │   └── defaults.yaml
│   │   │   ├── vlep/
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   └── defaults_cap_instruct.yaml
│   │   │   ├── vsr/
│   │   │   │   ├── defaults.yaml
│   │   │   │   ├── defaults_classification.yaml
│   │   │   │   ├── defaults_classification_instruct.yaml
│   │   │   │   └── defaults_instruct.yaml
│   │   │   ├── wavcaps/
│   │   │   │   ├── defaults_mm_cap.yaml
│   │   │   │   └── defaults_mm_cap_instruct.yaml
│   │   │   ├── webvid/
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   └── defaults_cap_instruct.yaml
│   │   │   ├── youcook/
│   │   │   │   ├── defaults_cap.yaml
│   │   │   │   └── defaults_cap_instruct.yaml
│   │   │   └── yt8m/
│   │   │       └── defaults_mm_dial.yaml
│   │   ├── default.yaml
│   │   └── models/
│   │       ├── albef_classification_ve.yaml
│   │       ├── albef_feature_extractor.yaml
│   │       ├── albef_nlvr.yaml
│   │       ├── albef_pretrain_base.yaml
│   │       ├── albef_retrieval_coco.yaml
│   │       ├── albef_retrieval_flickr.yaml
│   │       ├── albef_vqav2.yaml
│   │       ├── alpro_qa_msrvtt.yaml
│   │       ├── alpro_qa_msvd.yaml
│   │       ├── alpro_retrieval_didemo.yaml
│   │       ├── alpro_retrieval_msrvtt.yaml
│   │       ├── bert_config.json
│   │       ├── bert_config_alpro.json
│   │       ├── blip-diffusion/
│   │       │   ├── blip_diffusion_base.yaml
│   │       │   ├── blip_diffusion_controlnet_canny.yaml
│   │       │   ├── blip_diffusion_controlnet_depth.yaml
│   │       │   └── blip_diffusion_controlnet_hed.yaml
│   │       ├── blip2/
│   │       │   ├── blip2_caption_flant5xl.yaml
│   │       │   ├── blip2_caption_opt2.7b.yaml
│   │       │   ├── blip2_caption_opt6.7b.yaml
│   │       │   ├── blip2_coco.yaml
│   │       │   ├── blip2_instruct_flant5xl.yaml
│   │       │   ├── blip2_instruct_flant5xxl.yaml
│   │       │   ├── blip2_instruct_vicuna13b.yaml
│   │       │   ├── blip2_instruct_vicuna7b.yaml
│   │       │   ├── blip2_pretrain.yaml
│   │       │   ├── blip2_pretrain_flant5xl.yaml
│   │       │   ├── blip2_pretrain_flant5xl_vitL.yaml
│   │       │   ├── blip2_pretrain_flant5xxl.yaml
│   │       │   ├── blip2_pretrain_llama7b.yaml
│   │       │   ├── blip2_pretrain_opt2.7b.yaml
│   │       │   ├── blip2_pretrain_opt6.7b.yaml
│   │       │   ├── blip2_pretrain_vitL.yaml
│   │       │   ├── blip2_xinstruct_vicuna13b.yaml
│   │       │   └── blip2_xinstruct_vicuna7b.yaml
│   │       ├── blip_caption_base_coco.yaml
│   │       ├── blip_caption_large_coco.yaml
│   │       ├── blip_classification_base.yaml
│   │       ├── blip_feature_extractor_base.yaml
│   │       ├── blip_itm_base.yaml
│   │       ├── blip_itm_large.yaml
│   │       ├── blip_nlvr.yaml
│   │       ├── blip_pretrain_base.yaml
│   │       ├── blip_pretrain_large.yaml
│   │       ├── blip_retrieval_coco.yaml
│   │       ├── blip_retrieval_flickr.yaml
│   │       ├── blip_vqa_aokvqa.yaml
│   │       ├── blip_vqa_okvqa.yaml
│   │       ├── blip_vqav2.yaml
│   │       ├── clip/
│   │       │   ├── RN101-quickgelu.json
│   │       │   ├── RN101.json
│   │       │   ├── RN50-quickgelu.json
│   │       │   ├── RN50.json
│   │       │   ├── RN50x16.json
│   │       │   ├── RN50x4.json
│   │       │   ├── ViT-B-16-plus-240.json
│   │       │   ├── ViT-B-16-plus.json
│   │       │   ├── ViT-B-16.json
│   │       │   ├── ViT-B-32-plus-256.json
│   │       │   ├── ViT-B-32-quickgelu.json
│   │       │   ├── ViT-B-32.json
│   │       │   ├── ViT-H-14.json
│   │       │   ├── ViT-H-16.json
│   │       │   ├── ViT-L-14-280.json
│   │       │   ├── ViT-L-14-336.json
│   │       │   ├── ViT-L-14.json
│   │       │   ├── ViT-L-16-320.json
│   │       │   ├── ViT-L-16.json
│   │       │   ├── ViT-g-14.json
│   │       │   ├── timm-efficientnetv2_rw_s.json
│   │       │   ├── timm-resnet50d.json
│   │       │   ├── timm-resnetaa50d.json
│   │       │   ├── timm-resnetblur50.json
│   │       │   ├── timm-swin_base_patch4_window7_224.json
│   │       │   ├── timm-vit_base_patch16_224.json
│   │       │   ├── timm-vit_base_patch32_224.json
│   │       │   └── timm-vit_small_patch16_224.json
│   │       ├── clip_resnet50.yaml
│   │       ├── clip_vit_base16.yaml
│   │       ├── clip_vit_base32.yaml
│   │       ├── clip_vit_large14.yaml
│   │       ├── clip_vit_large14_336.yaml
│   │       ├── gpt_dialogue_base.yaml
│   │       ├── img2prompt-vqa/
│   │       │   └── img2prompt_vqa_base.yaml
│   │       ├── med_config.json
│   │       ├── med_config_albef.json
│   │       ├── med_large_config.json
│   │       └── pnp-vqa/
│   │           ├── pnp_vqa_3b.yaml
│   │           ├── pnp_vqa_base.yaml
│   │           ├── pnp_vqa_large.yaml
│   │           ├── unifiedqav2_3b_config.json
│   │           ├── unifiedqav2_base_config.json
│   │           └── unifiedqav2_large_config.json
│   ├── datasets/
│   │   ├── builders/
│   │   │   ├── __init__.py
│   │   │   ├── audio_caption_builder.py
│   │   │   ├── audio_qa_builder.py
│   │   │   ├── base_dataset_builder.py
│   │   │   ├── caption_builder.py
│   │   │   ├── classification_builder.py
│   │   │   ├── dialogue_builder.py
│   │   │   ├── discrn_builders.py
│   │   │   ├── image_text_pair_builder.py
│   │   │   ├── imagefolder_builder.py
│   │   │   ├── object3d_caption_builder.py
│   │   │   ├── object3d_classification_builder.py
│   │   │   ├── object3d_qa_builder.py
│   │   │   ├── retrieval_builder.py
│   │   │   ├── text_to_image_generation_builder.py
│   │   │   ├── video_qa_builder.py
│   │   │   └── vqa_builder.py
│   │   ├── data_utils.py
│   │   ├── datasets/
│   │   │   ├── aok_vqa_datasets.py
│   │   │   ├── audio_captioning_datasets.py
│   │   │   ├── audio_classification_datasets.py
│   │   │   ├── audio_qa_datasets.py
│   │   │   ├── avsd_dialogue_datasets.py
│   │   │   ├── base_dataset.py
│   │   │   ├── capfilt_dataset.py
│   │   │   ├── caption_datasets.py
│   │   │   ├── coco_caption_datasets.py
│   │   │   ├── coco_vqa_datasets.py
│   │   │   ├── dataloader_utils.py
│   │   │   ├── dialogue_datasets.py
│   │   │   ├── discriminatory_reasoning_datasets.py
│   │   │   ├── gqa_datasets.py
│   │   │   ├── iconqa_datasets.py
│   │   │   ├── image_text_pair_datasets.py
│   │   │   ├── imagefolder_dataset.py
│   │   │   ├── laion_dataset.py
│   │   │   ├── llava150k_dataset.py
│   │   │   ├── multimodal_classification_datasets.py
│   │   │   ├── music_avqa.py
│   │   │   ├── nlvr_datasets.py
│   │   │   ├── object3d_captioning_datasets.py
│   │   │   ├── object3d_classification_datasets.py
│   │   │   ├── object3d_qa_datasets.py
│   │   │   ├── ocr_datasets.py
│   │   │   ├── retrieval_datasets.py
│   │   │   ├── snli_ve_datasets.py
│   │   │   ├── subject_driven_t2i_dataset.py
│   │   │   ├── textcaps_datasets.py
│   │   │   ├── valor_caption.py
│   │   │   ├── vatex_captioning_datasets.py
│   │   │   ├── vg_vqa_datasets.py
│   │   │   ├── video_caption_datasets.py
│   │   │   ├── video_vqa_datasets.py
│   │   │   ├── violin_dataset.py
│   │   │   ├── visdial_dialogue_datasets.py
│   │   │   ├── vizwiz_vqa_datasets.py
│   │   │   ├── vlep_dataset.py
│   │   │   ├── vqa_datasets.py
│   │   │   ├── vsr_datasets.py
│   │   │   └── yt8m_video_dialogue_datasets.py
│   │   └── download_scripts/
│   │       ├── DownloadConceptualCaptions/
│   │       │   ├── LICENSE
│   │       │   ├── README.md
│   │       │   ├── create_annotation_12m.ipynb
│   │       │   ├── create_annotation_3m.ipynb
│   │       │   ├── download_data_cc12m.py
│   │       │   └── download_data_cc3m.py
│   │       ├── download_charade.py
│   │       ├── download_coco.py
│   │       ├── download_coin.py
│   │       ├── download_didemo.py
│   │       ├── download_flickr.py
│   │       ├── download_gqa.py
│   │       ├── download_iconqa.py
│   │       ├── download_msrvtt.py
│   │       ├── download_msvd.py
│   │       ├── download_nocaps.py
│   │       ├── download_sbu.py
│   │       ├── download_vg.py
│   │       └── download_violin.py
│   ├── models/
│   │   ├── __init__.py
│   │   ├── albef_models/
│   │   │   ├── __init__.py
│   │   │   ├── albef_classification.py
│   │   │   ├── albef_feature_extractor.py
│   │   │   ├── albef_nlvr.py
│   │   │   ├── albef_outputs.py
│   │   │   ├── albef_pretrain.py
│   │   │   ├── albef_retrieval.py
│   │   │   └── albef_vqa.py
│   │   ├── alpro_models/
│   │   │   ├── __init__.py
│   │   │   ├── alpro_outputs.py
│   │   │   ├── alpro_qa.py
│   │   │   └── alpro_retrieval.py
│   │   ├── base_model.py
│   │   ├── beats/
│   │   │   ├── BEATs.py
│   │   │   ├── LICENSE_BEATs.txt
│   │   │   ├── README.md
│   │   │   ├── Tokenizers.py
│   │   │   ├── backbone.py
│   │   │   ├── modules.py
│   │   │   └── quantizer.py
│   │   ├── beats_encoder.py
│   │   ├── blip2_models/
│   │   │   ├── Qformer.py
│   │   │   ├── __init__.py
│   │   │   ├── blip2.py
│   │   │   ├── blip2_image_text_matching.py
│   │   │   ├── blip2_opt.py
│   │   │   ├── blip2_qformer.py
│   │   │   ├── blip2_t5.py
│   │   │   ├── blip2_t5_instruct.py
│   │   │   ├── blip2_vicuna_instruct.py
│   │   │   ├── blip2_vicuna_xinstruct.py
│   │   │   ├── modeling_llama.py
│   │   │   ├── modeling_opt.py
│   │   │   └── modeling_t5.py
│   │   ├── blip_diffusion_models/
│   │   │   ├── __init__.py
│   │   │   ├── blip_diffusion.py
│   │   │   ├── modeling_ctx_clip.py
│   │   │   ├── ptp_utils.py
│   │   │   └── utils.py
│   │   ├── blip_models/
│   │   │   ├── __init__.py
│   │   │   ├── blip.py
│   │   │   ├── blip_caption.py
│   │   │   ├── blip_classification.py
│   │   │   ├── blip_feature_extractor.py
│   │   │   ├── blip_image_text_matching.py
│   │   │   ├── blip_nlvr.py
│   │   │   ├── blip_outputs.py
│   │   │   ├── blip_pretrain.py
│   │   │   ├── blip_retrieval.py
│   │   │   ├── blip_vqa.py
│   │   │   └── nlvr_encoder.py
│   │   ├── clip_models/
│   │   │   ├── __init__.py
│   │   │   ├── clip_outputs.py
│   │   │   ├── loss.py
│   │   │   ├── model.py
│   │   │   ├── pretrained.py
│   │   │   ├── timm_model.py
│   │   │   ├── tokenizer.py
│   │   │   ├── transform.py
│   │   │   └── utils.py
│   │   ├── clip_vit.py
│   │   ├── eva_vit.py
│   │   ├── gpt_models/
│   │   │   └── gpt_dialogue.py
│   │   ├── img2prompt_models/
│   │   │   ├── __init__.py
│   │   │   └── img2prompt_vqa.py
│   │   ├── med.py
│   │   ├── pnp_vqa_models/
│   │   │   ├── __init__.py
│   │   │   ├── pnp_unifiedqav2_fid.py
│   │   │   └── pnp_vqa.py
│   │   ├── timesformer/
│   │   │   ├── __init__.py
│   │   │   ├── conv2d_same.py
│   │   │   ├── features.py
│   │   │   ├── helpers.py
│   │   │   ├── linear.py
│   │   │   ├── vit.py
│   │   │   └── vit_utils.py
│   │   ├── ulip_models/
│   │   │   ├── ULIP_models.py
│   │   │   ├── losses.py
│   │   │   ├── pointbert/
│   │   │   │   ├── PointTransformer_8192point.yaml
│   │   │   │   ├── checkpoint.py
│   │   │   │   ├── dvae.py
│   │   │   │   ├── logger.py
│   │   │   │   ├── misc.py
│   │   │   │   └── point_encoder.py
│   │   │   ├── ulip_scaled_up_config.yaml
│   │   │   └── utils/
│   │   │       ├── __init__.py
│   │   │       ├── build.py
│   │   │       ├── config.py
│   │   │       ├── io.py
│   │   │       ├── logger.py
│   │   │       ├── registry.py
│   │   │       ├── tokenizer.py
│   │   │       └── utils.py
│   │   └── vit.py
│   ├── processors/
│   │   ├── __init__.py
│   │   ├── alpro_processors.py
│   │   ├── audio_processors.py
│   │   ├── base_processor.py
│   │   ├── blip_diffusion_processors.py
│   │   ├── blip_processors.py
│   │   ├── clip_processors.py
│   │   ├── functional_video.py
│   │   ├── gpt_processors.py
│   │   ├── instruction_text_processors.py
│   │   ├── randaugment.py
│   │   ├── transforms_video.py
│   │   └── ulip_processors.py
│   ├── projects/
│   │   ├── albef/
│   │   │   ├── eval/
│   │   │   │   ├── nlvr_eval.yaml
│   │   │   │   ├── ret_coco_eval.yaml
│   │   │   │   ├── ret_flickr30k_eval.yaml
│   │   │   │   ├── snli_ve_eval.yaml
│   │   │   │   ├── vqa_test.yaml
│   │   │   │   └── vqa_val.yaml
│   │   │   └── train/
│   │   │       ├── aokvqa_ft.yaml
│   │   │       ├── nlvr_ft.yaml
│   │   │       ├── okvqa_ft.yaml
│   │   │       ├── pretrain.yaml
│   │   │       ├── ret_coco_ft.yaml
│   │   │       ├── ret_flickr30k_ft.yaml
│   │   │       ├── snli_ve_ft.yaml
│   │   │       └── vqa_ft.yaml
│   │   ├── alpro/
│   │   │   ├── eval/
│   │   │   │   ├── didemo_ret_eval.yaml
│   │   │   │   ├── msrvtt_qa_eval.yaml
│   │   │   │   ├── msrvtt_ret_eval.yaml
│   │   │   │   └── msvd_qa_eval.yaml
│   │   │   └── train/
│   │   │       ├── didemo_ret_ft.yaml
│   │   │       ├── msrvtt_qa_ft.yaml
│   │   │       ├── msrvtt_retrieval_ft.yaml
│   │   │       └── msvd_qa_ft.yaml
│   │   ├── blip/
│   │   │   ├── coco_cap_ft_iter.yaml
│   │   │   ├── eval/
│   │   │   │   ├── aokvqa_eval.yaml
│   │   │   │   ├── caption_coco_eval.yaml
│   │   │   │   ├── caption_coco_eval_large.yaml
│   │   │   │   ├── nlvr_eval.yaml
│   │   │   │   ├── nocaps_eval.yaml
│   │   │   │   ├── okvqa_eval.yaml
│   │   │   │   ├── ret_coco_eval.yaml
│   │   │   │   ├── ret_flickr_eval.yaml
│   │   │   │   └── vqav2_eval.yaml
│   │   │   └── train/
│   │   │       ├── aokvqa_ft.yaml
│   │   │       ├── caption_coco_ft.yaml
│   │   │       ├── caption_coco_large_ft.yaml
│   │   │       ├── nlvr_ft.yaml
│   │   │       ├── okvqa_ft.yaml
│   │   │       ├── pretrain_14m.yaml
│   │   │       ├── retrieval_coco_ft.yaml
│   │   │       ├── retrieval_flickr_ft.yaml
│   │   │       └── vqav2_ft.yaml
│   │   ├── blip2/
│   │   │   ├── eval/
│   │   │   │   ├── caption_coco_flant5xl_eval.yaml
│   │   │   │   ├── caption_coco_opt2.7b_eval.yaml
│   │   │   │   ├── caption_coco_opt6.7b_eval.yaml
│   │   │   │   ├── caption_nocaps_out_domain_flant5xl_eval.yaml
│   │   │   │   ├── caption_nocaps_out_domain_flant5xxl_eval.yaml
│   │   │   │   ├── gqa_zeroshot_flant5xl_eval.yaml
│   │   │   │   ├── okvqa_zeroshot_flant5xl_eval.yaml
│   │   │   │   ├── ret_coco_eval.yaml
│   │   │   │   ├── ret_flickr_eval.yaml
│   │   │   │   ├── vqav2_zeroshot_flant5xl_eval.yaml
│   │   │   │   └── vqav2_zeroshot_opt_eval.yaml
│   │   │   └── train/
│   │   │       ├── caption_coco_ft.yaml
│   │   │       ├── pretrain_stage1.yaml
│   │   │       ├── pretrain_stage2.yaml
│   │   │       └── retrieval_coco_ft.yaml
│   │   ├── blip_diffusion/
│   │   │   ├── finetune-db-dog.yaml
│   │   │   ├── finetune-db-pink-dress.yaml
│   │   │   ├── finetune-db-shein-jacket.yaml
│   │   │   └── finetune-db-template.yaml
│   │   ├── clip/
│   │   │   ├── exp_coco_ret_eval.yaml
│   │   │   ├── exp_flickr_ret_eval.yaml
│   │   │   └── exp_imnet_zs_eval.yaml
│   │   ├── gpt/
│   │   │   ├── eval/
│   │   │   │   └── dialogue_avsd_eval.yaml
│   │   │   └── train/
│   │   │       └── dialogue_avsd_ft.yaml
│   │   ├── instructblip/
│   │   │   ├── caption_coco_flant5xl_eval_test.yaml
│   │   │   ├── caption_coco_flant5xl_eval_val.yaml
│   │   │   ├── caption_coco_flant5xxl_eval_test.yaml
│   │   │   ├── caption_coco_flant5xxl_eval_val.yaml
│   │   │   ├── caption_coco_vicuna13b_eval_test.yaml
│   │   │   ├── caption_coco_vicuna13b_eval_val.yaml
│   │   │   ├── caption_coco_vicuna7b_eval_test.yaml
│   │   │   ├── caption_coco_vicuna7b_eval_val.yaml
│   │   │   ├── caption_msrvtt_flant5xl_eval_test.yaml
│   │   │   ├── caption_msrvtt_flant5xl_eval_val.yaml
│   │   │   ├── caption_msrvtt_flant5xxl_eval_test.yaml
│   │   │   ├── caption_msrvtt_flant5xxl_eval_val.yaml
│   │   │   ├── caption_msrvtt_vicuna13b_eval_test.yaml
│   │   │   ├── caption_msrvtt_vicuna13b_eval_val.yaml
│   │   │   ├── caption_msrvtt_vicuna7b_eval_test.yaml
│   │   │   ├── caption_msrvtt_vicuna7b_eval_val.yaml
│   │   │   ├── caption_msvd_flant5xl_eval.yaml
│   │   │   ├── caption_msvd_flant5xxl_eval.yaml
│   │   │   ├── caption_msvd_vicuna13b_eval.yaml
│   │   │   ├── caption_msvd_vicuna7b_eval.yaml
│   │   │   ├── caption_nocaps_out_domain_flant5xl_eval.yaml
│   │   │   ├── caption_nocaps_out_domain_flant5xxl_eval.yaml
│   │   │   ├── caption_nocaps_out_domain_vicuna13b_eval.yaml
│   │   │   ├── caption_nocaps_out_domain_vicuna7b_eval.yaml
│   │   │   ├── caption_vatex_flant5xl_eval.yaml
│   │   │   ├── caption_vatex_flant5xxl_eval.yaml
│   │   │   ├── caption_vatex_vicuna13b_eval.yaml
│   │   │   ├── caption_vatex_vicuna7b_eval.yaml
│   │   │   ├── classification_modelnet40_vicuna13b.yaml
│   │   │   ├── classification_modelnet40_vicuna7b.yaml
│   │   │   ├── classification_snlive_flant5xl.yaml
│   │   │   ├── classification_snlive_flant5xxl.yaml
│   │   │   ├── classification_snlive_vicuna13b.yaml
│   │   │   ├── classification_snlive_vicuna13b_test.yaml
│   │   │   ├── classification_snlive_vicuna7b_test.yaml
│   │   │   ├── classification_snlive_vicuna7b_val.yaml
│   │   │   ├── completion_modelnet40_vicuna13b.yaml
│   │   │   ├── completion_modelnet40_vicuna7b.yaml
│   │   │   ├── qa_msrvtt_flant5xl_eval_test.yaml
│   │   │   ├── qa_msrvtt_flant5xxl_eval_test.yaml
│   │   │   ├── qa_msrvtt_vicuna13b_eval_test.yaml
│   │   │   ├── qa_msrvtt_vicuna7b_eval_test.yaml
│   │   │   ├── qa_msvd_flant5xl_eval.yaml
│   │   │   ├── qa_msvd_flant5xxl_eval.yaml
│   │   │   ├── qa_msvd_vicuna13b_eval.yaml
│   │   │   ├── qa_msvd_vicuna7b_eval.yaml
│   │   │   ├── qa_okvqa_flant5xl_eval.yaml
│   │   │   ├── qa_okvqa_flant5xxl_eval.yaml
│   │   │   ├── qa_okvqa_vicuna13b_eval.yaml
│   │   │   └── qa_okvqa_vicuna7b_eval.yaml
│   │   ├── pnp-vqa/
│   │   │   └── eval/
│   │   │       ├── gqa_eval.yaml
│   │   │       ├── gqa_eval_3b.yaml
│   │   │       ├── gqa_eval_large.yaml
│   │   │       ├── okvqa_eval.yaml
│   │   │       ├── okvqa_eval_3b.yaml
│   │   │       ├── okvqa_eval_large.yaml
│   │   │       ├── vqav2_eval.yaml
│   │   │       ├── vqav2_eval_3b.yaml
│   │   │       ├── vqav2_eval_large.yaml
│   │   │       ├── vqav2_test_eval.yaml
│   │   │       ├── vqav2_test_eval_3b.yaml
│   │   │       └── vqav2_test_eval_large.yaml
│   │   └── xinstruct_blip/
│   │       ├── eval/
│   │       │   ├── discrn/
│   │       │   │   ├── audio_video_caption.yaml
│   │       │   │   ├── audio_video_caption_13b.yaml
│   │       │   │   ├── audio_video_describe.yaml
│   │       │   │   ├── audio_video_describe_13b.yaml
│   │       │   │   ├── audio_video_describe_nocue.yaml
│   │       │   │   ├── audio_video_describe_proj copy.yaml
│   │       │   │   ├── audio_video_describe_proj.yaml
│   │       │   │   ├── audio_video_describe_rand_init.yaml
│   │       │   │   ├── image_3d_caption.yaml
│   │       │   │   ├── image_3d_caption_13b.yaml
│   │       │   │   ├── image_3d_describe.yaml
│   │       │   │   ├── image_3d_describe_13b.yaml
│   │       │   │   ├── image_3d_describe_no_init.yaml
│   │       │   │   ├── image_3d_describe_nocue.yaml
│   │       │   │   └── image_3d_describe_proj.yaml
│   │       │   ├── vicuna13b/
│   │       │   │   ├── audio/
│   │       │   │   │   ├── audiocaps_captioning_qa.yaml
│   │       │   │   │   ├── audiocaps_captioning_test.yaml
│   │       │   │   │   ├── audiocaps_captioning_val.yaml
│   │       │   │   │   ├── clothoQA_captioning.yaml
│   │       │   │   │   ├── clothov1_captioning.yaml
│   │       │   │   │   ├── clothov2_captioning.yaml
│   │       │   │   │   ├── esc50_classification.yaml
│   │       │   │   │   └── esc50_classification_completion.yaml
│   │       │   │   ├── crossmodal/
│   │       │   │   │   ├── musicavqa/
│   │       │   │   │   │   ├── musicavqa_audio_eval.yaml
│   │       │   │   │   │   ├── musicavqa_joint_eval.yaml
│   │       │   │   │   │   └── musicavqa_video_eval.yaml
│   │       │   │   │   └── vatex/
│   │       │   │   │       ├── vatex_audio_captioning.yaml
│   │       │   │   │       ├── vatex_captioning.yaml
│   │       │   │   │       ├── vatex_joint_captioning.yaml
│   │       │   │   │       └── vatex_joint_captioning_interleave.yaml
│   │       │   │   ├── image/
│   │       │   │   │   ├── coco_captioning_test.yaml
│   │       │   │   │   ├── coco_captioning_val.yaml
│   │       │   │   │   ├── flickr30k_captioning.yaml
│   │       │   │   │   ├── gqa_qa.yaml
│   │       │   │   │   ├── nocaps_captioning.yaml
│   │       │   │   │   ├── nocaps_out_domain_captioning.yaml
│   │       │   │   │   ├── okvqa_qa.yaml
│   │       │   │   │   ├── snlive_classification_test.yaml
│   │       │   │   │   ├── snlive_classification_val.yaml
│   │       │   │   │   └── vizwiz_qa.yaml
│   │       │   │   ├── image_with_coco/
│   │       │   │   │   ├── coco_captioning_test.yaml
│   │       │   │   │   ├── coco_captioning_val.yaml
│   │       │   │   │   ├── flickr30k_captioning.yaml
│   │       │   │   │   ├── gqa_qa.yaml
│   │       │   │   │   ├── nocaps_captioning.yaml
│   │       │   │   │   ├── nocaps_out_domain_captioning.yaml
│   │       │   │   │   ├── okvqa_qa.yaml
│   │       │   │   │   ├── snlive_classification_test.yaml
│   │       │   │   │   ├── snlive_classification_val.yaml
│   │       │   │   │   └── vizwiz_qa.yaml
│   │       │   │   ├── pc/
│   │       │   │   │   ├── modelnet40_classification.yaml
│   │       │   │   │   ├── modelnet40_completion.yaml
│   │       │   │   │   ├── objaverse_captioning.yaml
│   │       │   │   │   └── objaverse_qa.yaml
│   │       │   │   ├── video/
│   │       │   │   │   ├── msrvtt_captioning.yaml
│   │       │   │   │   ├── msrvtt_captioning_test.yaml
│   │       │   │   │   ├── msrvtt_captioning_val.yaml
│   │       │   │   │   ├── msrvtt_qa_test.yaml
│   │       │   │   │   ├── msrvtt_qa_val.yaml
│   │       │   │   │   ├── msvd_captioning.yaml
│   │       │   │   │   ├── msvd_qa.yaml
│   │       │   │   │   ├── vatex_audio_captioning.yaml
│   │       │   │   │   ├── vatex_captioning.yaml
│   │       │   │   │   ├── vatex_joint_captioning.yaml
│   │       │   │   │   └── vatex_joint_captioning_interleave.yaml
│   │       │   │   └── video_image/
│   │       │   │       ├── msvd_captioning.yaml
│   │       │   │       ├── msvd_qa.yaml
│   │       │   │       └── vatex_captioning.yaml
│   │       │   ├── vicuna7b/
│   │       │   │   ├── audio/
│   │       │   │   │   ├── audiocaps_captioning_qa.yaml
│   │       │   │   │   ├── audiocaps_captioning_test.yaml
│   │       │   │   │   ├── audiocaps_captioning_val.yaml
│   │       │   │   │   ├── clothoQA_captioning.yaml
│   │       │   │   │   ├── clothov1_captioning.yaml
│   │       │   │   │   ├── clothov2_captioning.yaml
│   │       │   │   │   ├── esc50_classification.yaml
│   │       │   │   │   └── esc50_classification_completion.yaml
│   │       │   │   ├── audio_no_init/
│   │       │   │   │   ├── audiocaps_captioning_qa.yaml
│   │       │   │   │   ├── audiocaps_captioning_test.yaml
│   │       │   │   │   ├── audiocaps_captioning_val.yaml
│   │       │   │   │   ├── clothoQA_captioning.yaml
│   │       │   │   │   ├── clothov1_captioning.yaml
│   │       │   │   │   ├── clothov2_captioning.yaml
│   │       │   │   │   ├── esc50_classification.yaml
│   │       │   │   │   └── esc50_classification_completion.yaml
│   │       │   │   ├── audio_projection_only/
│   │       │   │   │   ├── audiocaps_captioning_qa.yaml
│   │       │   │   │   ├── audiocaps_captioning_test.yaml
│   │       │   │   │   ├── audiocaps_captioning_val.yaml
│   │       │   │   │   ├── clothoQA_captioning.yaml
│   │       │   │   │   ├── clothov1_captioning.yaml
│   │       │   │   │   ├── clothov2_captioning.yaml
│   │       │   │   │   ├── esc50_classification.yaml
│   │       │   │   │   └── esc50_classification_completion.yaml
│   │       │   │   ├── audio_projection_only_nocue/
│   │       │   │   │   ├── audiocaps_captioning_qa.yaml
│   │       │   │   │   ├── audiocaps_captioning_test.yaml
│   │       │   │   │   ├── audiocaps_captioning_val.yaml
│   │       │   │   │   ├── clothoQA_captioning.yaml
│   │       │   │   │   ├── clothov1_captioning.yaml
│   │       │   │   │   ├── clothov2_captioning.yaml
│   │       │   │   │   ├── esc50_classification.yaml
│   │       │   │   │   └── esc50_classification_completion.yaml
│   │       │   │   ├── crossmodal/
│   │       │   │   │   ├── musicavqa/
│   │       │   │   │   │   ├── musicavqa_audio_eval.yaml
│   │       │   │   │   │   ├── musicavqa_joint_eval.yaml
│   │       │   │   │   │   └── musicavqa_video_eval.yaml
│   │       │   │   │   └── vatex/
│   │       │   │   │       ├── vatex_audio_captioning.yaml
│   │       │   │   │       ├── vatex_captioning.yaml
│   │       │   │   │       ├── vatex_joint_captioning.yaml
│   │       │   │   │       └── vatex_joint_captioning_interleave.yaml
│   │       │   │   ├── image/
│   │       │   │   │   ├── coco_captioning_test.yaml
│   │       │   │   │   ├── coco_captioning_val.yaml
│   │       │   │   │   ├── flickr30k_captioning.yaml
│   │       │   │   │   ├── gqa_qa.yaml
│   │       │   │   │   ├── gqa_qa_val.yaml
│   │       │   │   │   ├── nocaps_captioning.yaml
│   │       │   │   │   ├── nocaps_out_domain_captioning.yaml
│   │       │   │   │   ├── okvqa_qa.yaml
│   │       │   │   │   ├── snlive_classification_test.yaml
│   │       │   │   │   ├── snlive_classification_val.yaml
│   │       │   │   │   └── vizwiz_qa.yaml
│   │       │   │   ├── image_full_init/
│   │       │   │   │   ├── coco_captioning_test.yaml
│   │       │   │   │   ├── coco_captioning_val.yaml
│   │       │   │   │   ├── flickr30k_captioning.yaml
│   │       │   │   │   ├── gqa_qa.yaml
│   │       │   │   │   ├── gqa_qa_val.yaml
│   │       │   │   │   ├── nocaps_captioning.yaml
│   │       │   │   │   ├── nocaps_out_domain_captioning.yaml
│   │       │   │   │   ├── okvqa_qa.yaml
│   │       │   │   │   ├── snlive_classification_test.yaml
│   │       │   │   │   ├── snlive_classification_val.yaml
│   │       │   │   │   └── vizwiz_qa.yaml
│   │       │   │   ├── image_no_init/
│   │       │   │   │   ├── coco_captioning_test.yaml
│   │       │   │   │   ├── coco_captioning_val.yaml
│   │       │   │   │   ├── flickr30k_captioning.yaml
│   │       │   │   │   ├── gqa_qa.yaml
│   │       │   │   │   ├── gqa_qa_val.yaml
│   │       │   │   │   ├── nocaps_captioning.yaml
│   │       │   │   │   ├── nocaps_out_domain_captioning.yaml
│   │       │   │   │   ├── okvqa_qa.yaml
│   │       │   │   │   ├── snlive_classification_test.yaml
│   │       │   │   │   ├── snlive_classification_val.yaml
│   │       │   │   │   └── vizwiz_qa.yaml
│   │       │   │   ├── image_pre_coco/
│   │       │   │   │   ├── coco_captioning_test.yaml
│   │       │   │   │   ├── coco_captioning_val.yaml
│   │       │   │   │   ├── flickr30k_captioning.yaml
│   │       │   │   │   ├── gqa_qa.yaml
│   │       │   │   │   ├── nocaps_captioning.yaml
│   │       │   │   │   ├── nocaps_out_domain_captioning.yaml
│   │       │   │   │   ├── okvqa_qa.yaml
│   │       │   │   │   ├── snlive_classification_test.yaml
│   │       │   │   │   ├── snlive_classification_val.yaml
│   │       │   │   │   └── vizwiz_qa.yaml
│   │       │   │   ├── image_projection_only/
│   │       │   │   │   ├── coco_captioning_test.yaml
│   │       │   │   │   ├── coco_captioning_val.yaml
│   │       │   │   │   ├── flickr30k_captioning.yaml
│   │       │   │   │   ├── gqa_qa.yaml
│   │       │   │   │   ├── gqa_qa_val.yaml
│   │       │   │   │   ├── nocaps_captioning.yaml
│   │       │   │   │   ├── nocaps_out_domain_captioning.yaml
│   │       │   │   │   ├── okvqa_qa.yaml
│   │       │   │   │   ├── snlive_classification_test.yaml
│   │       │   │   │   ├── snlive_classification_val.yaml
│   │       │   │   │   └── vizwiz_qa.yaml
│   │       │   │   ├── pc/
│   │       │   │   │   ├── modelnet40_classification.yaml
│   │       │   │   │   ├── modelnet40_completion.yaml
│   │       │   │   │   ├── objaverse_captioning.yaml
│   │       │   │   │   └── objaverse_qa.yaml
│   │       │   │   ├── pc_no_init/
│   │       │   │   │   ├── modelnet40_classification.yaml
│   │       │   │   │   ├── modelnet40_completion.yaml
│   │       │   │   │   ├── objaverse_captioning.yaml
│   │       │   │   │   └── objaverse_qa.yaml
│   │       │   │   ├── pc_projection_only/
│   │       │   │   │   ├── modelnet40_classification.yaml
│   │       │   │   │   ├── modelnet40_completion.yaml
│   │       │   │   │   ├── objaverse_captioning.yaml
│   │       │   │   │   └── objaverse_qa.yaml
│   │       │   │   ├── pc_ulip1/
│   │       │   │   │   ├── modelnet40_classification.yaml
│   │       │   │   │   ├── modelnet40_completion.yaml
│   │       │   │   │   ├── objaverse_captioning.yaml
│   │       │   │   │   └── objaverse_qa.yaml
│   │       │   │   ├── pc_ulip2_scaled_up/
│   │       │   │   │   ├── modelnet40_classification.yaml
│   │       │   │   │   ├── modelnet40_completion.yaml
│   │       │   │   │   ├── objaverse_captioning.yaml
│   │       │   │   │   └── objaverse_qa.yaml
│   │       │   │   ├── pc_ulip_objaverse/
│   │       │   │   │   ├── modelnet40_classification.yaml
│   │       │   │   │   ├── modelnet40_completion.yaml
│   │       │   │   │   ├── objaverse_captioning.yaml
│   │       │   │   │   └── objaverse_qa.yaml
│   │       │   │   ├── pc_ulip_objaverse_shapenet/
│   │       │   │   │   ├── modelnet40_classification.yaml
│   │       │   │   │   ├── modelnet40_completion.yaml
│   │       │   │   │   ├── objaverse_captioning.yaml
│   │       │   │   │   └── objaverse_qa.yaml
│   │       │   │   ├── pc_ulip_shapenet/
│   │       │   │   │   ├── modelnet40_classification.yaml
│   │       │   │   │   ├── modelnet40_completion.yaml
│   │       │   │   │   ├── objaverse_captioning.yaml
│   │       │   │   │   └── objaverse_qa.yaml
│   │       │   │   ├── video/
│   │       │   │   │   ├── msrvtt_captioning_test.yaml
│   │       │   │   │   ├── msrvtt_captioning_val.yaml
│   │       │   │   │   ├── msrvtt_qa_test.yaml
│   │       │   │   │   ├── msrvtt_qa_val.yaml
│   │       │   │   │   ├── msvd_captioning.yaml
│   │       │   │   │   ├── msvd_qa.yaml
│   │       │   │   │   └── vatex_captioning.yaml
│   │       │   │   ├── video_image/
│   │       │   │   │   ├── msvd_captioning.yaml
│   │       │   │   │   ├── msvd_qa.yaml
│   │       │   │   │   └── vatex_captioning.yaml
│   │       │   │   ├── video_image_pre_coco/
│   │       │   │   │   ├── msvd_captioning.yaml
│   │       │   │   │   ├── msvd_qa.yaml
│   │       │   │   │   └── vatex_captioning.yaml
│   │       │   │   └── video_no_upsample/
│   │       │   │       ├── msrvtt_captioning_test.yaml
│   │       │   │       ├── msrvtt_captioning_val.yaml
│   │       │   │       ├── msrvtt_qa_test.yaml
│   │       │   │       ├── msrvtt_qa_val.yaml
│   │       │   │       ├── msvd_captioning.yaml
│   │       │   │       ├── msvd_captioning_up.yaml
│   │       │   │       ├── msvd_qa.yaml
│   │       │   │       ├── msvd_qa_up.yaml
│   │       │   │       ├── vatex_captioning.yaml
│   │       │   │       └── vatex_captioning_up.yaml
│   │       │   └── vicuna7b_nocue/
│   │       │       ├── audio/
│   │       │       │   ├── audiocaps_captioning_qa.yaml
│   │       │       │   ├── audiocaps_captioning_test.yaml
│   │       │       │   ├── audiocaps_captioning_val.yaml
│   │       │       │   ├── clothoQA_captioning.yaml
│   │       │       │   ├── clothov1_captioning.yaml
│   │       │       │   ├── clothov2_captioning.yaml
│   │       │       │   ├── esc50_classification.yaml
│   │       │       │   └── esc50_classification_completion.yaml
│   │       │       ├── crossmodal/
│   │       │       │   ├── musicavqa/
│   │       │       │   │   ├── musicavqa_audio_eval.yaml
│   │       │       │   │   ├── musicavqa_joint_eval.yaml
│   │       │       │   │   └── musicavqa_video_eval.yaml
│   │       │       │   └── vatex/
│   │       │       │       ├── vatex_audio_captioning.yaml
│   │       │       │       ├── vatex_captioning.yaml
│   │       │       │       └── vatex_joint_captioning.yaml
│   │       │       ├── image/
│   │       │       │   ├── coco_captioning_test.yaml
│   │       │       │   ├── coco_captioning_val.yaml
│   │       │       │   ├── flickr30k_captioning.yaml
│   │       │       │   ├── gqa_qa.yaml
│   │       │       │   ├── nocaps_captioning.yaml
│   │       │       │   ├── nocaps_out_domain_captioning.yaml
│   │       │       │   ├── okvqa_qa.yaml
│   │       │       │   ├── snlive_classification_test.yaml
│   │       │       │   ├── snlive_classification_val.yaml
│   │       │       │   └── vizwiz_qa.yaml
│   │       │       ├── pc/
│   │       │       │   ├── modelnet40_classification.yaml
│   │       │       │   ├── modelnet40_completion.yaml
│   │       │       │   ├── objaverse_captioning.yaml
│   │       │       │   └── objaverse_qa.yaml
│   │       │       ├── video/
│   │       │       │   ├── msrvtt_captioning_test.yaml
│   │       │       │   ├── msrvtt_captioning_val.yaml
│   │       │       │   ├── msrvtt_qa_test.yaml
│   │       │       │   ├── msrvtt_qa_val.yaml
│   │       │       │   ├── msvd_captioning.yaml
│   │       │       │   ├── msvd_qa.yaml
│   │       │       │   └── vatex_captioning.yaml
│   │       │       └── video_image/
│   │       │           ├── msvd_captioning.yaml
│   │       │           ├── msvd_qa.yaml
│   │       │           └── vatex_captioning.yaml
│   │       ├── prompt_variation/
│   │       │   └── nocaps/
│   │       │       ├── instructblip/
│   │       │       │   ├── original.yaml
│   │       │       │   ├── template_1.yaml
│   │       │       │   ├── template_2.yaml
│   │       │       │   ├── template_3.yaml
│   │       │       │   ├── template_4.yaml
│   │       │       │   └── template_5.yaml
│   │       │       └── xinstructblip/
│   │       │           ├── template_1.yaml
│   │       │           ├── template_2.yaml
│   │       │           ├── template_3.yaml
│   │       │           ├── template_4.yaml
│   │       │           └── template_5.yaml
│   │       └── train/
│   │           ├── vicuna13b/
│   │           │   ├── audio_training.yaml
│   │           │   ├── audio_training_continue.yaml
│   │           │   ├── image_train.yaml
│   │           │   ├── image_train_continue.yaml
│   │           │   ├── pc_training.yaml
│   │           │   └── video_training.yaml
│   │           ├── vicuna7b/
│   │           │   ├── audio_training.yaml
│   │           │   ├── audio_training_improved.yaml
│   │           │   ├── audio_training_no_init.yaml
│   │           │   ├── audio_training_projection_only.yaml
│   │           │   ├── audio_training_projection_only_nocue.yaml
│   │           │   ├── image_train.yaml
│   │           │   ├── image_train_improved.yaml
│   │           │   ├── image_train_no_init.yaml
│   │           │   ├── image_train_projection_only.yaml
│   │           │   ├── lora_training.yaml
│   │           │   ├── pc_training.yaml
│   │           │   ├── pc_training_improved.yaml
│   │           │   ├── pc_training_no_init.yaml
│   │           │   ├── pc_training_projection_only.yaml
│   │           │   ├── pc_training_projection_only_nocue.yaml
│   │           │   ├── pc_training_scaled_up.yaml
│   │           │   ├── pc_training_ulip1.yaml
│   │           │   ├── pc_training_ulip2_objaverse_shapenet_k_1.yaml
│   │           │   ├── pc_training_ulip_objaverse.yaml
│   │           │   ├── pc_training_ulip_shapenet.yaml
│   │           │   ├── video_training.yaml
│   │           │   └── video_training_no_msrvtt_upsample.yaml
│   │           └── vicuna7b_nocue/
│   │               ├── audio_training.yaml
│   │               ├── image_train.yaml
│   │               ├── pc_training.yaml
│   │               └── video_training.yaml
│   ├── runners/
│   │   ├── __init__.py
│   │   ├── runner_base.py
│   │   └── runner_iter.py
│   └── tasks/
│       ├── __init__.py
│       ├── base_task.py
│       ├── captioning.py
│       ├── dialogue.py
│       ├── image_text_pretrain.py
│       ├── multimodal_classification.py
│       ├── retrieval.py
│       ├── text_to_image_generation.py
│       ├── vqa.py
│       └── vqa_reading_comprehension.py
├── projects/
│   ├── blip-diffusion/
│   │   ├── README.md
│   │   └── notebooks/
│   │       ├── editing_real_finetuned.ipynb
│   │       ├── editing_real_zeroshot.ipynb
│   │       ├── editing_synthetic_zeroshot.ipynb
│   │       ├── editing_tryon_zeroshot.ipynb
│   │       ├── generation_finetuned_dog.ipynb
│   │       ├── generation_zeroshot.ipynb
│   │       └── stylization.ipynb
│   ├── blip2/
│   │   └── README.md
│   ├── img2llm-vqa/
│   │   ├── README.md
│   │   ├── img2llm_vqa.ipynb
│   │   └── img2llm_vqa.py
│   ├── img2prompt-vqa/
│   │   └── README.md
│   ├── instructblip/
│   │   ├── README.md
│   │   └── run_demo.py
│   ├── pnp-vqa/
│   │   ├── README.md
│   │   └── pnp_vqa.ipynb
│   └── xinstructblip/
│       ├── README.md
│       ├── data_aug/
│       │   ├── 3d_qa_data_generation.py
│       │   └── audio_qa_data_generation.py
│       ├── demo/
│       │   ├── configs/
│       │   │   ├── vicuna13b.yaml
│       │   │   ├── vicuna7b.yaml
│       │   │   ├── vicuna7b_blip_init.yaml
│       │   │   ├── vicuna7b_no_init.yaml
│       │   │   ├── vicuna7b_nocue.yaml
│       │   │   ├── vicuna7b_projection.yaml
│       │   │   ├── vicuna7b_rand.yaml
│       │   │   └── vicuna7b_v2.yaml
│       │   ├── demo.ipynb
│       │   ├── examples/
│       │   │   └── point_cloud/
│       │   │       └── banana.glb
│       │   └── run_demo.py
│       ├── discrn/
│       │   ├── caption_baseline/
│       │   │   ├── predict_audio.py
│       │   │   ├── predict_image.py
│       │   │   ├── predict_pc.py
│       │   │   ├── predict_video.py
│       │   │   └── render_images.py
│       │   └── data_generation/
│       │       ├── audiocaps_video_audio.py
│       │       └── objaverse_img_3d.py
│       └── modelnet_baseline/
│           └── render_images.py
├── pyproject.toml
├── requirements.txt
├── run_scripts/
│   ├── albef/
│   │   ├── eval/
│   │   │   ├── eval_albef_nlvr.sh
│   │   │   ├── eval_albef_ve.sh
│   │   │   ├── eval_coco_retrieval.sh
│   │   │   ├── eval_flickr30k_retrieval.sh
│   │   │   ├── test_albef_vqa.sh
│   │   │   └── val_albef_vqa.sh
│   │   └── train/
│   │       ├── pretrain.sh
│   │       ├── train_aokvqa_albef.sh
│   │       ├── train_coco_retrieval_albef.sh
│   │       ├── train_flickr30k_retrieval_albef.sh
│   │       ├── train_nlvr_albef.sh
│   │       ├── train_okvqa_albef.sh
│   │       ├── train_ve_albef.sh
│   │       └── train_vqa_albef.sh
│   ├── alpro/
│   │   ├── eval/
│   │   │   ├── eval_didemo_ret.sh
│   │   │   ├── eval_msrvtt_qa.sh
│   │   │   ├── eval_msrvtt_ret.sh
│   │   │   └── eval_msvd_qa.sh
│   │   └── train/
│   │       ├── train_didemo_ret.sh
│   │       ├── train_msrvtt_qa.sh
│   │       ├── train_msrvtt_ret.sh
│   │       └── train_msvd_qa.sh
│   ├── blip/
│   │   ├── eval/
│   │   │   ├── eval_aokvqa.sh
│   │   │   ├── eval_coco_cap.sh
│   │   │   ├── eval_coco_cap_large.sh
│   │   │   ├── eval_nlvr.sh
│   │   │   ├── eval_nocaps.sh
│   │   │   ├── eval_okvqa.sh
│   │   │   ├── eval_ret_coco.sh
│   │   │   ├── eval_ret_flickr.sh
│   │   │   └── validate_vqa.sh
│   │   └── train/
│   │       ├── pretrain.sh
│   │       ├── train_aokvqa.sh
│   │       ├── train_caption_coco.sh
│   │       ├── train_caption_coco_large.sh
│   │       ├── train_caption_coco_large_iters.sh
│   │       ├── train_nlvr.sh
│   │       ├── train_okvqa.sh
│   │       ├── train_retrieval_coco.sh
│   │       ├── train_retrieval_flickr.sh
│   │       └── train_vqa.sh
│   ├── blip-diffusion/
│   │   ├── train_db.sh
│   │   ├── train_db_dog.sh
│   │   ├── train_db_jacket_s.sh
│   │   ├── train_db_pink_dress.sh
│   │   └── train_db_shein_jacket.sh
│   ├── blip2/
│   │   ├── eval/
│   │   │   ├── eval_cap_coco_flant5xl.sh
│   │   │   ├── eval_cap_coco_opt2.7b.sh
│   │   │   ├── eval_cap_coco_opt6.7b.sh
│   │   │   ├── eval_gqa_zeroshot_flant5xl.sh
│   │   │   ├── eval_okvqa_zeroshot_flant5xl.sh
│   │   │   ├── eval_ret_coco.sh
│   │   │   ├── eval_ret_flickr.sh
│   │   │   ├── validate_vqa_zeroshot_flant5xl.sh
│   │   │   └── validate_vqa_zeroshot_opt.sh
│   │   └── train/
│   │       ├── pretrain_stage1.sh
│   │       ├── pretrain_stage2.sh
│   │       ├── train_caption_coco.sh
│   │       └── train_retrieval_coco.sh
│   ├── clip/
│   │   └── eval/
│   │       ├── eval_clip_ret_coco.sh
│   │       ├── eval_clip_ret_flickr.sh
│   │       └── eval_clip_zs_imnet.sh
│   ├── gpt/
│   │   ├── eval/
│   │   │   └── eval_video_dialogue_avsd.sh
│   │   └── train/
│   │       └── train_video_dialogue_avsd.sh
│   ├── pnp-vqa/
│   │   └── eval/
│   │       ├── eval_gqa.sh
│   │       ├── eval_gqa_3b.sh
│   │       ├── eval_gqa_large.sh
│   │       ├── eval_okvqa.sh
│   │       ├── eval_okvqa_3b.sh
│   │       ├── eval_okvqa_large.sh
│   │       ├── eval_vqav2.sh
│   │       ├── eval_vqav2_3b.sh
│   │       ├── eval_vqav2_large.sh
│   │       ├── eval_vqav2_test.sh
│   │       ├── eval_vqav2_test_3b.sh
│   │       └── eval_vqav2_test_large.sh
│   ├── run_browser.sh
│   └── run_demo.sh
├── setup.py
├── tests/
│   └── models/
│       ├── test_albef.py
│       ├── test_blip.py
│       ├── test_blip2.py
│       └── test_pnp_vqa.py
└── train.py

SYMBOL INDEX (4547 symbols across 475 files)

FILE: app/__init__.py
  function load_demo_image (line 16) | def load_demo_image():

FILE: app/calculate_coco_features.py
  function load_demo_image (line 22) | def load_demo_image():
  function read_img (line 31) | def read_img(filepath):

FILE: app/caption.py
  function app (line 15) | def app():
  function generate_caption (line 72) | def generate_caption(

FILE: app/classification.py
  function load_demo_image (line 23) | def load_demo_image(img_url=None):
  function load_model_cache (line 38) | def load_model_cache(model_type, device):
  function app (line 63) | def app():

FILE: app/dataset_browser.py
  function sample_dataset (line 26) | def sample_dataset(dataset, indices):
  function get_concat_v (line 32) | def get_concat_v(im1, im2):
  function resize_img_w (line 43) | def resize_img_w(raw_img, new_w=224):
  function get_visual_key (line 58) | def get_visual_key(dataset):
  function gather_items (line 69) | def gather_items(samples, exclude=[]):
  function load_dataset_cache (line 84) | def load_dataset_cache(name):
  function format_text (line 88) | def format_text(text):
  function show_samples (line 94) | def show_samples(dataset, offset=0, is_next=False):

FILE: app/image_text_match.py
  function app (line 19) | def app():

FILE: app/multimodal_search.py
  function load_feat (line 34) | def load_feat():
  function load_feature_extractor_model (line 61) | def load_feature_extractor_model(device):
  function app (line 72) | def app():
  function read_and_process_images (line 183) | def read_and_process_images(image_paths, vis_processor):
  function compute_gradcam_batch (line 191) | def compute_gradcam_batch(model, visual_input, text_input, tokenized_tex...

FILE: app/multipage.py
  class MultiPage (line 17) | class MultiPage:
    method __init__ (line 20) | def __init__(self) -> None:
    method add_page (line 24) | def add_page(self, title, func) -> None:
    method run (line 34) | def run(self):

FILE: app/text_localization.py
  function app (line 20) | def app():

FILE: app/utils.py
  function resize_img (line 18) | def resize_img(raw_img):
  function read_img (line 25) | def read_img(filepath):
  function load_model_cache (line 39) | def load_model_cache(name, model_type, is_eval, device):
  function init_bert_tokenizer (line 44) | def init_bert_tokenizer():
  function getAttMap (line 49) | def getAttMap(img, attMap, blur=True, overlap=True):
  function load_blip_itm_model (line 77) | def load_blip_itm_model(device, model_type="base"):

FILE: app/vqa.py
  function app (line 15) | def app():

FILE: evaluate.py
  function parse_args (line 33) | def parse_args():
  function setup_seeds (line 52) | def setup_seeds(config):
  function main (line 63) | def main():

FILE: lavis/common/annotator/canny/__init__.py
  class CannyDetector (line 4) | class CannyDetector:
    method __call__ (line 5) | def __call__(self, img, low_threshold, high_threshold):

FILE: lavis/common/annotator/hed/__init__.py
  class Network (line 9) | class Network(torch.nn.Module):
    method __init__ (line 10) | def __init__(self, model_path):
    method forward (line 71) | def forward(self, tenInput):
  class HEDdetector (line 96) | class HEDdetector:
    method __init__ (line 97) | def __init__(self):
    method __call__ (line 105) | def __call__(self, input_image):
  function nms (line 117) | def nms(x, t, s):

FILE: lavis/common/annotator/midas/__init__.py
  class MidasDetector (line 9) | class MidasDetector:
    method __init__ (line 10) | def __init__(self):
    method __call__ (line 13) | def __call__(self, input_image, a=np.pi * 2.0, bg_th=0.1):

FILE: lavis/common/annotator/midas/api.py
  function disabled_train (line 26) | def disabled_train(self, mode=True):
  function load_midas_transform (line 32) | def load_midas_transform(model_type):
  function load_model (line 77) | def load_model(model_type):
  class MiDaSInference (line 145) | class MiDaSInference(nn.Module):
    method __init__ (line 158) | def __init__(self, model_type):
    method forward (line 165) | def forward(self, x):

FILE: lavis/common/annotator/midas/midas/base_model.py
  class BaseModel (line 4) | class BaseModel(torch.nn.Module):
    method load (line 5) | def load(self, path):

FILE: lavis/common/annotator/midas/midas/blocks.py
  function _make_encoder (line 11) | def _make_encoder(backbone, features, use_pretrained, groups=1, expand=F...
  function _make_scratch (line 49) | def _make_scratch(in_shape, out_shape, groups=1, expand=False):
  function _make_pretrained_efficientnet_lite3 (line 78) | def _make_pretrained_efficientnet_lite3(use_pretrained, exportable=False):
  function _make_efficientnet_backbone (line 88) | def _make_efficientnet_backbone(effnet):
  function _make_resnet_backbone (line 101) | def _make_resnet_backbone(resnet):
  function _make_pretrained_resnext101_wsl (line 114) | def _make_pretrained_resnext101_wsl(use_pretrained):
  class Interpolate (line 120) | class Interpolate(nn.Module):
    method __init__ (line 124) | def __init__(self, scale_factor, mode, align_corners=False):
    method forward (line 138) | def forward(self, x):
  class ResidualConvUnit (line 155) | class ResidualConvUnit(nn.Module):
    method __init__ (line 159) | def __init__(self, features):
    method forward (line 177) | def forward(self, x):
  class FeatureFusionBlock (line 194) | class FeatureFusionBlock(nn.Module):
    method __init__ (line 198) | def __init__(self, features):
    method forward (line 209) | def forward(self, *xs):
  class ResidualConvUnit_custom (line 231) | class ResidualConvUnit_custom(nn.Module):
    method __init__ (line 235) | def __init__(self, features, activation, bn):
    method forward (line 263) | def forward(self, x):
  class FeatureFusionBlock_custom (line 291) | class FeatureFusionBlock_custom(nn.Module):
    method __init__ (line 295) | def __init__(self, features, activation, deconv=False, bn=False, expan...
    method forward (line 320) | def forward(self, *xs):

FILE: lavis/common/annotator/midas/midas/dpt_depth.py
  function _make_fusion_block (line 15) | def _make_fusion_block(features, use_bn):
  class DPT (line 26) | class DPT(BaseModel):
    method __init__ (line 27) | def __init__(
    method forward (line 67) | def forward(self, x):
  class DPTDepthModel (line 88) | class DPTDepthModel(DPT):
    method __init__ (line 89) | def __init__(self, path=None, non_negative=True, **kwargs):
    method forward (line 107) | def forward(self, x):

FILE: lavis/common/annotator/midas/midas/midas_net.py
  class MidasNet (line 12) | class MidasNet(BaseModel):
    method __init__ (line 16) | def __init__(self, path=None, features=256, non_negative=True):
    method forward (line 49) | def forward(self, x):

FILE: lavis/common/annotator/midas/midas/midas_net_custom.py
  class MidasNet_small (line 12) | class MidasNet_small(BaseModel):
    method __init__ (line 16) | def __init__(self, path=None, features=64, backbone="efficientnet_lite...
    method forward (line 73) | def forward(self, x):
  function fuse_model (line 109) | def fuse_model(m):

FILE: lavis/common/annotator/midas/midas/transforms.py
  function apply_min_size (line 6) | def apply_min_size(sample, size, image_interpolation_method=cv2.INTER_AR...
  class Resize (line 48) | class Resize(object):
    method __init__ (line 52) | def __init__(
    method constrain_to_multiple_of (line 94) | def constrain_to_multiple_of(self, x, min_val=0, max_val=None):
    method get_size (line 105) | def get_size(self, width, height):
    method __call__ (line 162) | def __call__(self, sample):
  class NormalizeImage (line 197) | class NormalizeImage(object):
    method __init__ (line 201) | def __init__(self, mean, std):
    method __call__ (line 205) | def __call__(self, sample):
  class PrepareForNet (line 211) | class PrepareForNet(object):
    method __init__ (line 215) | def __init__(self):
    method __call__ (line 218) | def __call__(self, sample):

FILE: lavis/common/annotator/midas/midas/vit.py
  class Slice (line 9) | class Slice(nn.Module):
    method __init__ (line 10) | def __init__(self, start_index=1):
    method forward (line 14) | def forward(self, x):
  class AddReadout (line 18) | class AddReadout(nn.Module):
    method __init__ (line 19) | def __init__(self, start_index=1):
    method forward (line 23) | def forward(self, x):
  class ProjectReadout (line 31) | class ProjectReadout(nn.Module):
    method __init__ (line 32) | def __init__(self, in_features, start_index=1):
    method forward (line 38) | def forward(self, x):
  class Transpose (line 45) | class Transpose(nn.Module):
    method __init__ (line 46) | def __init__(self, dim0, dim1):
    method forward (line 51) | def forward(self, x):
  function forward_vit (line 56) | def forward_vit(pretrained, x):
  function _resize_pos_embed (line 100) | def _resize_pos_embed(self, posemb, gs_h, gs_w):
  function forward_flex (line 117) | def forward_flex(self, x):
  function get_activation (line 159) | def get_activation(name):
  function get_readout_oper (line 166) | def get_readout_oper(vit_features, features, use_readout, start_index=1):
  function _make_vit_b16_backbone (line 183) | def _make_vit_b16_backbone(
  function _make_pretrained_vitl16_384 (line 297) | def _make_pretrained_vitl16_384(pretrained, use_readout="ignore", hooks=...
  function _make_pretrained_vitb16_384 (line 310) | def _make_pretrained_vitb16_384(pretrained, use_readout="ignore", hooks=...
  function _make_pretrained_deitb16_384 (line 319) | def _make_pretrained_deitb16_384(pretrained, use_readout="ignore", hooks...
  function _make_pretrained_deitb16_distil_384 (line 328) | def _make_pretrained_deitb16_distil_384(pretrained, use_readout="ignore"...
  function _make_vit_b_rn50_backbone (line 343) | def _make_vit_b_rn50_backbone(
  function _make_pretrained_vitb_rn50_384 (line 478) | def _make_pretrained_vitb_rn50_384(

FILE: lavis/common/annotator/midas/utils.py
  function read_pfm (line 9) | def read_pfm(path):
  function write_pfm (line 58) | def write_pfm(path, image, scale=1):
  function read_image (line 97) | def read_image(path):
  function resize_image (line 116) | def resize_image(img):
  function resize_depth (line 146) | def resize_depth(depth, width, height):
  function write_depth (line 165) | def write_depth(path, depth, bits=1):

FILE: lavis/common/annotator/mlsd/__init__.py
  class MLSDdetector (line 17) | class MLSDdetector:
    method __init__ (line 18) | def __init__(self):
    method __call__ (line 27) | def __call__(self, input_image, thr_v, thr_d):

FILE: lavis/common/annotator/mlsd/models/mbv2_mlsd_large.py
  class BlockTypeA (line 9) | class BlockTypeA(nn.Module):
    method __init__ (line 10) | def __init__(self, in_c1, in_c2, out_c1, out_c2, upscale = True):
    method forward (line 24) | def forward(self, a, b):
  class BlockTypeB (line 32) | class BlockTypeB(nn.Module):
    method __init__ (line 33) | def __init__(self, in_c, out_c):
    method forward (line 46) | def forward(self, x):
  class BlockTypeC (line 51) | class BlockTypeC(nn.Module):
    method __init__ (line 52) | def __init__(self, in_c, out_c):
    method forward (line 66) | def forward(self, x):
  function _make_divisible (line 72) | def _make_divisible(v, divisor, min_value=None):
  class ConvBNReLU (line 92) | class ConvBNReLU(nn.Sequential):
    method __init__ (line 93) | def __init__(self, in_planes, out_planes, kernel_size=3, stride=1, gro...
    method forward (line 112) | def forward(self, x):
  class InvertedResidual (line 124) | class InvertedResidual(nn.Module):
    method __init__ (line 125) | def __init__(self, inp, oup, stride, expand_ratio):
    method forward (line 146) | def forward(self, x):
  class MobileNetV2 (line 153) | class MobileNetV2(nn.Module):
    method __init__ (line 154) | def __init__(self, pretrained=True):
    method _forward_impl (line 218) | def _forward_impl(self, x):
    method forward (line 233) | def forward(self, x):
    method _load_pretrained_model (line 236) | def _load_pretrained_model(self):
  class MobileV2_MLSD_Large (line 247) | class MobileV2_MLSD_Large(nn.Module):
    method __init__ (line 248) | def __init__(self):
    method forward (line 275) | def forward(self, x):

FILE: lavis/common/annotator/mlsd/models/mbv2_mlsd_tiny.py
  class BlockTypeA (line 9) | class BlockTypeA(nn.Module):
    method __init__ (line 10) | def __init__(self, in_c1, in_c2, out_c1, out_c2, upscale = True):
    method forward (line 24) | def forward(self, a, b):
  class BlockTypeB (line 31) | class BlockTypeB(nn.Module):
    method __init__ (line 32) | def __init__(self, in_c, out_c):
    method forward (line 45) | def forward(self, x):
  class BlockTypeC (line 50) | class BlockTypeC(nn.Module):
    method __init__ (line 51) | def __init__(self, in_c, out_c):
    method forward (line 65) | def forward(self, x):
  function _make_divisible (line 71) | def _make_divisible(v, divisor, min_value=None):
  class ConvBNReLU (line 91) | class ConvBNReLU(nn.Sequential):
    method __init__ (line 92) | def __init__(self, in_planes, out_planes, kernel_size=3, stride=1, gro...
    method forward (line 111) | def forward(self, x):
  class InvertedResidual (line 123) | class InvertedResidual(nn.Module):
    method __init__ (line 124) | def __init__(self, inp, oup, stride, expand_ratio):
    method forward (line 145) | def forward(self, x):
  class MobileNetV2 (line 152) | class MobileNetV2(nn.Module):
    method __init__ (line 153) | def __init__(self, pretrained=True):
    method _forward_impl (line 218) | def _forward_impl(self, x):
    method forward (line 233) | def forward(self, x):
    method _load_pretrained_model (line 236) | def _load_pretrained_model(self):
  class MobileV2_MLSD_Tiny (line 247) | class MobileV2_MLSD_Tiny(nn.Module):
    method __init__ (line 248) | def __init__(self):
    method forward (line 263) | def forward(self, x):

FILE: lavis/common/annotator/mlsd/utils.py
  function deccode_output_score_and_ptss (line 19) | def deccode_output_score_and_ptss(tpMap, topk_n = 200, ksize = 5):
  function pred_lines (line 47) | def pred_lines(image, model,
  function pred_squares (line 89) | def pred_squares(image,

FILE: lavis/common/annotator/openpose/__init__.py
  class OpenposeDetector (line 16) | class OpenposeDetector:
    method __init__ (line 17) | def __init__(self):
    method __call__ (line 29) | def __call__(self, oriImg, hand=False):

FILE: lavis/common/annotator/openpose/body.py
  class Body (line 14) | class Body(object):
    method __init__ (line 15) | def __init__(self, model_path):
    method __call__ (line 24) | def __call__(self, oriImg):

FILE: lavis/common/annotator/openpose/hand.py
  class Hand (line 15) | class Hand(object):
    method __init__ (line 16) | def __init__(self, model_path):
    method __call__ (line 25) | def __call__(self, oriImg):

FILE: lavis/common/annotator/openpose/model.py
  function make_layers (line 7) | def make_layers(block, no_relu_layers):
  class bodypose_model (line 24) | class bodypose_model(nn.Module):
    method __init__ (line 25) | def __init__(self):
    method forward (line 114) | def forward(self, x):
  class handpose_model (line 143) | class handpose_model(nn.Module):
    method __init__ (line 144) | def __init__(self):
    method forward (line 204) | def forward(self, x):

FILE: lavis/common/annotator/openpose/util.py
  function padRightDownCorner (line 7) | def padRightDownCorner(img, stride, padValue):
  function transfer (line 30) | def transfer(model, model_weights):
  function draw_bodypose (line 37) | def draw_bodypose(canvas, candidate, subset):
  function draw_handpose (line 74) | def draw_handpose(canvas, all_hand_peaks, show_number=False):
  function handDetect (line 94) | def handDetect(candidate, subset, oriImg):
  function npmax (line 159) | def npmax(array):

FILE: lavis/common/annotator/uniformer/__init__.py
  class UniformerDetector (line 11) | class UniformerDetector:
    method __init__ (line 12) | def __init__(self):
    method __call__ (line 20) | def __call__(self, img):

FILE: lavis/common/annotator/uniformer/mmcv/arraymisc/quantization.py
  function quantize (line 5) | def quantize(arr, min_val, max_val, levels, dtype=np.int64):
  function dequantize (line 32) | def dequantize(arr, min_val, max_val, levels, dtype=np.float64):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/alexnet.py
  class AlexNet (line 7) | class AlexNet(nn.Module):
    method __init__ (line 14) | def __init__(self, num_classes=-1):
    method init_weights (line 43) | def init_weights(self, pretrained=None):
    method forward (line 54) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/activation.py
  class Clamp (line 18) | class Clamp(nn.Module):
    method __init__ (line 31) | def __init__(self, min=-1., max=1.):
    method forward (line 36) | def forward(self, x):
  class GELU (line 48) | class GELU(nn.Module):
    method forward (line 70) | def forward(self, input):
  function build_activation_layer (line 81) | def build_activation_layer(cfg):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/context_block.py
  function last_zero_init (line 9) | def last_zero_init(m):
  class ContextBlock (line 17) | class ContextBlock(nn.Module):
    method __init__ (line 36) | def __init__(self,
    method reset_parameters (line 75) | def reset_parameters(self):
    method spatial_pool (line 85) | def spatial_pool(self, x):
    method forward (line 111) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/conv.py
  function build_conv_layer (line 12) | def build_conv_layer(cfg, *args, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/conv2d_adaptive_padding.py
  class Conv2dAdaptivePadding (line 11) | class Conv2dAdaptivePadding(nn.Conv2d):
    method __init__ (line 33) | def __init__(self,
    method forward (line 45) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/conv_module.py
  class ConvModule (line 16) | class ConvModule(nn.Module):
    method __init__ (line 70) | def __init__(self,
    method norm (line 169) | def norm(self):
    method init_weights (line 175) | def init_weights(self):
    method forward (line 196) | def forward(self, x, activate=True, norm=True):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/conv_ws.py
  function conv_ws_2d (line 9) | def conv_ws_2d(input,
  class ConvWS2d (line 26) | class ConvWS2d(nn.Conv2d):
    method __init__ (line 28) | def __init__(self,
    method forward (line 49) | def forward(self, x):
  class ConvAWS2d (line 55) | class ConvAWS2d(nn.Conv2d):
    method __init__ (line 78) | def __init__(self,
    method _get_weight (line 101) | def _get_weight(self, weight):
    method forward (line 109) | def forward(self, x):
    method _load_from_state_dict (line 114) | def _load_from_state_dict(self, state_dict, prefix, local_metadata, st...

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/depthwise_separable_conv_module.py
  class DepthwiseSeparableConvModule (line 7) | class DepthwiseSeparableConvModule(nn.Module):
    method __init__ (line 48) | def __init__(self,
    method forward (line 93) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/drop.py
  function drop_path (line 9) | def drop_path(x, drop_prob=0., training=False):
  class DropPath (line 28) | class DropPath(nn.Module):
    method __init__ (line 39) | def __init__(self, drop_prob=0.1):
    method forward (line 43) | def forward(self, x):
  class Dropout (line 48) | class Dropout(nn.Dropout):
    method __init__ (line 59) | def __init__(self, drop_prob=0.5, inplace=False):
  function build_dropout (line 63) | def build_dropout(cfg, default_args=None):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/generalized_attention.py
  class GeneralizedAttention (line 14) | class GeneralizedAttention(nn.Module):
    method __init__ (line 47) | def __init__(self,
    method get_position_embedding (line 166) | def get_position_embedding(self,
    method forward (line 216) | def forward(self, x_input):
    method init_weights (line 403) | def init_weights(self):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/hsigmoid.py
  class HSigmoid (line 8) | class HSigmoid(nn.Module):
    method __init__ (line 23) | def __init__(self, bias=1.0, divisor=2.0, min_value=0.0, max_value=1.0):
    method forward (line 31) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/hswish.py
  class HSwish (line 8) | class HSwish(nn.Module):
    method __init__ (line 24) | def __init__(self, inplace=False):
    method forward (line 28) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/non_local.py
  class _NonLocalNd (line 12) | class _NonLocalNd(nn.Module, metaclass=ABCMeta):
    method __init__ (line 35) | def __init__(self,
    method init_weights (line 99) | def init_weights(self, std=0.01, zeros_init=True):
    method gaussian (line 116) | def gaussian(self, theta_x, phi_x):
    method embedded_gaussian (line 124) | def embedded_gaussian(self, theta_x, phi_x):
    method dot_product (line 135) | def dot_product(self, theta_x, phi_x):
    method concatenation (line 143) | def concatenation(self, theta_x, phi_x):
    method forward (line 160) | def forward(self, x):
  class NonLocal1d (line 214) | class NonLocal1d(_NonLocalNd):
    method __init__ (line 226) | def __init__(self,
  class NonLocal2d (line 246) | class NonLocal2d(_NonLocalNd):
    method __init__ (line 260) | def __init__(self,
  class NonLocal3d (line 279) | class NonLocal3d(_NonLocalNd):
    method __init__ (line 291) | def __init__(self,

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/norm.py
  function infer_abbr (line 23) | def infer_abbr(class_type):
  function build_norm_layer (line 72) | def build_norm_layer(cfg, num_features, postfix=''):
  function is_norm (line 122) | def is_norm(layer, exclude=None):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/padding.py
  function build_padding_layer (line 11) | def build_padding_layer(cfg, *args, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/plugin.py
  function infer_abbr (line 12) | def infer_abbr(class_type):
  function build_plugin_layer (line 55) | def build_plugin_layer(cfg, postfix='', **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/scale.py
  class Scale (line 6) | class Scale(nn.Module):
    method __init__ (line 16) | def __init__(self, scale=1.0):
    method forward (line 20) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/swish.py
  class Swish (line 9) | class Swish(nn.Module):
    method __init__ (line 21) | def __init__(self):
    method forward (line 24) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/transformer.py
  function build_positional_encoding (line 33) | def build_positional_encoding(cfg, default_args=None):
  function build_attention (line 38) | def build_attention(cfg, default_args=None):
  function build_feedforward_network (line 43) | def build_feedforward_network(cfg, default_args=None):
  function build_transformer_layer (line 48) | def build_transformer_layer(cfg, default_args=None):
  function build_transformer_layer_sequence (line 53) | def build_transformer_layer_sequence(cfg, default_args=None):
  class MultiheadAttention (line 59) | class MultiheadAttention(BaseModule):
    method __init__ (line 81) | def __init__(self,
    method forward (line 112) | def forward(self,
  class FFN (line 206) | class FFN(BaseModule):
    method __init__ (line 234) | def __init__(self,
    method forward (line 269) | def forward(self, x, identity=None):
  class BaseTransformerLayer (line 283) | class BaseTransformerLayer(BaseModule):
    method __init__ (line 319) | def __init__(self,
    method forward (line 412) | def forward(self,
  class TransformerLayerSequence (line 514) | class TransformerLayerSequence(BaseModule):
    method __init__ (line 533) | def __init__(self, transformerlayers=None, num_layers=None, init_cfg=N...
    method forward (line 549) | def forward(self,

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/upsample.py
  class PixelShufflePack (line 13) | class PixelShufflePack(nn.Module):
    method __init__ (line 27) | def __init__(self, in_channels, out_channels, scale_factor,
    method init_weights (line 41) | def init_weights(self):
    method forward (line 44) | def forward(self, x):
  function build_upsample_layer (line 50) | def build_upsample_layer(cfg, *args, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/bricks/wrappers.py
  function obsolete_torch_version (line 24) | def obsolete_torch_version(torch_version, version_threshold):
  class NewEmptyTensorOp (line 28) | class NewEmptyTensorOp(torch.autograd.Function):
    method forward (line 31) | def forward(ctx, x, new_shape):
    method backward (line 36) | def backward(ctx, grad):
  class Conv2d (line 42) | class Conv2d(nn.Conv2d):
    method forward (line 44) | def forward(self, x):
  class Conv3d (line 63) | class Conv3d(nn.Conv3d):
    method forward (line 65) | def forward(self, x):
  class ConvTranspose2d (line 86) | class ConvTranspose2d(nn.ConvTranspose2d):
    method forward (line 88) | def forward(self, x):
  class ConvTranspose3d (line 109) | class ConvTranspose3d(nn.ConvTranspose3d):
    method forward (line 111) | def forward(self, x):
  class MaxPool2d (line 129) | class MaxPool2d(nn.MaxPool2d):
    method forward (line 131) | def forward(self, x):
  class MaxPool3d (line 147) | class MaxPool3d(nn.MaxPool3d):
    method forward (line 149) | def forward(self, x):
  class Linear (line 166) | class Linear(torch.nn.Linear):
    method forward (line 168) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/builder.py
  function build_model_from_cfg (line 6) | def build_model_from_cfg(cfg, registry, default_args=None):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/resnet.py
  function conv3x3 (line 10) | def conv3x3(in_planes, out_planes, stride=1, dilation=1):
  class BasicBlock (line 22) | class BasicBlock(nn.Module):
    method __init__ (line 25) | def __init__(self,
    method forward (line 45) | def forward(self, x):
  class Bottleneck (line 64) | class Bottleneck(nn.Module):
    method __init__ (line 67) | def __init__(self,
    method forward (line 110) | def forward(self, x):
  function make_res_layer (line 143) | def make_res_layer(block,
  class ResNet (line 181) | class ResNet(nn.Module):
    method __init__ (line 210) | def __init__(self,
    method init_weights (line 265) | def init_weights(self, pretrained=None):
    method forward (line 279) | def forward(self, x):
    method train (line 295) | def train(self, mode=True):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/utils/flops_counter.py
  function get_model_complexity_info (line 36) | def get_model_complexity_info(model,
  function flops_to_string (line 118) | def flops_to_string(flops, units='GFLOPs', precision=2):
  function params_to_string (line 161) | def params_to_string(num_params, units=None, precision=2):
  function print_model_with_flops (line 198) | def print_model_with_flops(model,
  function get_model_parameters_number (line 307) | def get_model_parameters_number(model):
  function add_flops_counting_methods (line 320) | def add_flops_counting_methods(net_main_module):
  function compute_average_flops_cost (line 337) | def compute_average_flops_cost(self):
  function start_flops_count (line 355) | def start_flops_count(self):
  function stop_flops_count (line 378) | def stop_flops_count(self):
  function reset_flops_count (line 389) | def reset_flops_count(self):
  function empty_flops_counter_hook (line 400) | def empty_flops_counter_hook(module, input, output):
  function upsample_flops_counter_hook (line 404) | def upsample_flops_counter_hook(module, input, output):
  function relu_flops_counter_hook (line 413) | def relu_flops_counter_hook(module, input, output):
  function linear_flops_counter_hook (line 418) | def linear_flops_counter_hook(module, input, output):
  function pool_flops_counter_hook (line 425) | def pool_flops_counter_hook(module, input, output):
  function norm_flops_counter_hook (line 430) | def norm_flops_counter_hook(module, input, output):
  function deconv_flops_counter_hook (line 440) | def deconv_flops_counter_hook(conv_module, input, output):
  function conv_flops_counter_hook (line 467) | def conv_flops_counter_hook(conv_module, input, output):
  function batch_counter_hook (line 498) | def batch_counter_hook(module, input, output):
  function add_batch_counter_variables_or_reset (line 511) | def add_batch_counter_variables_or_reset(module):
  function add_batch_counter_hook_function (line 516) | def add_batch_counter_hook_function(module):
  function remove_batch_counter_hook_function (line 524) | def remove_batch_counter_hook_function(module):
  function add_flops_counter_variable_or_reset (line 530) | def add_flops_counter_variable_or_reset(module):
  function is_supported_instance (line 540) | def is_supported_instance(module):
  function remove_flops_counter_hook_function (line 546) | def remove_flops_counter_hook_function(module):
  function get_modules_mapping (line 553) | def get_modules_mapping():

FILE: lavis/common/annotator/uniformer/mmcv/cnn/utils/fuse_conv_bn.py
  function _fuse_conv_bn (line 6) | def _fuse_conv_bn(conv, bn):
  function fuse_conv_bn (line 27) | def fuse_conv_bn(module):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/utils/sync_bn.py
  class _BatchNormXd (line 6) | class _BatchNormXd(torch.nn.modules.batchnorm._BatchNorm):
    method _check_input_dim (line 17) | def _check_input_dim(self, input):
  function revert_sync_batchnorm (line 21) | def revert_sync_batchnorm(module):

FILE: lavis/common/annotator/uniformer/mmcv/cnn/utils/weight_init.py
  function update_init_info (line 16) | def update_init_info(module, init_info):
  function constant_init (line 48) | def constant_init(module, val, bias=0):
  function xavier_init (line 55) | def xavier_init(module, gain=1, bias=0, distribution='normal'):
  function normal_init (line 66) | def normal_init(module, mean=0, std=1, bias=0):
  function trunc_normal_init (line 73) | def trunc_normal_init(module: nn.Module,
  function uniform_init (line 85) | def uniform_init(module, a=0, b=1, bias=0):
  function kaiming_init (line 92) | def kaiming_init(module,
  function caffe2_xavier_init (line 110) | def caffe2_xavier_init(module, bias=0):
  function bias_init_with_prob (line 122) | def bias_init_with_prob(prior_prob):
  function _get_bases_name (line 128) | def _get_bases_name(m):
  class BaseInit (line 132) | class BaseInit(object):
    method __init__ (line 134) | def __init__(self, *, bias=0, bias_prob=None, layer=None):
    method _get_init_info (line 157) | def _get_init_info(self):
  class ConstantInit (line 163) | class ConstantInit(BaseInit):
    method __init__ (line 175) | def __init__(self, val, **kwargs):
    method __call__ (line 179) | def __call__(self, module):
    method _get_init_info (line 194) | def _get_init_info(self):
  class XavierInit (line 200) | class XavierInit(BaseInit):
    method __init__ (line 217) | def __init__(self, gain=1, distribution='normal', **kwargs):
    method __call__ (line 222) | def __call__(self, module):
    method _get_init_info (line 237) | def _get_init_info(self):
  class NormalInit (line 244) | class NormalInit(BaseInit):
    method __init__ (line 260) | def __init__(self, mean=0, std=1, **kwargs):
    method __call__ (line 265) | def __call__(self, module):
    method _get_init_info (line 280) | def _get_init_info(self):
  class TruncNormalInit (line 287) | class TruncNormalInit(BaseInit):
    method __init__ (line 306) | def __init__(self,
    method __call__ (line 318) | def __call__(self, module: nn.Module) -> None:
    method _get_init_info (line 335) | def _get_init_info(self):
  class UniformInit (line 342) | class UniformInit(BaseInit):
    method __init__ (line 358) | def __init__(self, a=0, b=1, **kwargs):
    method __call__ (line 363) | def __call__(self, module):
    method _get_init_info (line 378) | def _get_init_info(self):
  class KaimingInit (line 385) | class KaimingInit(BaseInit):
    method __init__ (line 411) | def __init__(self,
    method __call__ (line 423) | def __call__(self, module):
    method _get_init_info (line 440) | def _get_init_info(self):
  class Caffe2XavierInit (line 448) | class Caffe2XavierInit(KaimingInit):
    method __init__ (line 451) | def __init__(self, **kwargs):
    method __call__ (line 459) | def __call__(self, module):
  class PretrainedInit (line 464) | class PretrainedInit(object):
    method __init__ (line 478) | def __init__(self, checkpoint, prefix=None, map_location=None):
    method __call__ (line 483) | def __call__(self, module):
    method _get_init_info (line 506) | def _get_init_info(self):
  function _initialize (line 511) | def _initialize(module, cfg, wholemodule=False):
  function _initialize_override (line 520) | def _initialize_override(module, override, cfg):
  function initialize (line 550) | def initialize(module, init_cfg):
  function _no_grad_trunc_normal_ (line 622) | def _no_grad_trunc_normal_(tensor: Tensor, mean: float, std: float, a: f...
  function trunc_normal_ (line 662) | def trunc_normal_(tensor: Tensor,

FILE: lavis/common/annotator/uniformer/mmcv/cnn/vgg.py
  function conv3x3 (line 9) | def conv3x3(in_planes, out_planes, dilation=1):
  function make_vgg_layer (line 19) | def make_vgg_layer(inplanes,
  class VGG (line 37) | class VGG(nn.Module):
    method __init__ (line 61) | def __init__(self,
    method init_weights (line 125) | def init_weights(self, pretrained=None):
    method forward (line 141) | def forward(self, x):
    method train (line 159) | def train(self, mode=True):

FILE: lavis/common/annotator/uniformer/mmcv/engine/test.py
  function single_gpu_test (line 15) | def single_gpu_test(model, data_loader):
  function multi_gpu_test (line 44) | def multi_gpu_test(model, data_loader, tmpdir=None, gpu_collect=False):
  function collect_results_cpu (line 91) | def collect_results_cpu(result_part, size, tmpdir=None):
  function collect_results_gpu (line 155) | def collect_results_gpu(result_part, size):

FILE: lavis/common/annotator/uniformer/mmcv/fileio/file_client.py
  class BaseStorageBackend (line 19) | class BaseStorageBackend(metaclass=ABCMeta):
    method name (line 31) | def name(self):
    method allow_symlink (line 35) | def allow_symlink(self):
    method get (line 39) | def get(self, filepath):
    method get_text (line 43) | def get_text(self, filepath):
  class CephBackend (line 47) | class CephBackend(BaseStorageBackend):
    method __init__ (line 60) | def __init__(self, path_mapping=None):
    method get (line 72) | def get(self, filepath):
    method get_text (line 81) | def get_text(self, filepath, encoding=None):
  class PetrelBackend (line 85) | class PetrelBackend(BaseStorageBackend):
    method __init__ (line 108) | def __init__(self,
    method _map_path (line 121) | def _map_path(self, filepath: Union[str, Path]) -> str:
    method _format_path (line 134) | def _format_path(self, filepath: str) -> str:
    method get (line 147) | def get(self, filepath: Union[str, Path]) -> memoryview:
    method get_text (line 164) | def get_text(self,
    method put (line 179) | def put(self, obj: bytes, filepath: Union[str, Path]) -> None:
    method put_text (line 190) | def put_text(self,
    method remove (line 204) | def remove(self, filepath: Union[str, Path]) -> None:
    method exists (line 220) | def exists(self, filepath: Union[str, Path]) -> bool:
    method isdir (line 240) | def isdir(self, filepath: Union[str, Path]) -> bool:
    method isfile (line 261) | def isfile(self, filepath: Union[str, Path]) -> bool:
    method join_path (line 281) | def join_path(self, filepath: Union[str, Path],
    method get_local_path (line 300) | def get_local_path(self, filepath: Union[str, Path]) -> Iterable[str]:
    method list_dir_or_file (line 331) | def list_dir_or_file(self,
  class MemcachedBackend (line 413) | class MemcachedBackend(BaseStorageBackend):
    method __init__ (line 423) | def __init__(self, server_list_cfg, client_cfg, sys_path=None):
    method get (line 440) | def get(self, filepath):
    method get_text (line 447) | def get_text(self, filepath, encoding=None):
  class LmdbBackend (line 451) | class LmdbBackend(BaseStorageBackend):
    method __init__ (line 469) | def __init__(self,
    method get (line 488) | def get(self, filepath):
    method get_text (line 499) | def get_text(self, filepath, encoding=None):
  class HardDiskBackend (line 503) | class HardDiskBackend(BaseStorageBackend):
    method get (line 508) | def get(self, filepath: Union[str, Path]) -> bytes:
    method get_text (line 521) | def get_text(self,
    method put (line 538) | def put(self, obj: bytes, filepath: Union[str, Path]) -> None:
    method put_text (line 553) | def put_text(self,
    method remove (line 573) | def remove(self, filepath: Union[str, Path]) -> None:
    method exists (line 581) | def exists(self, filepath: Union[str, Path]) -> bool:
    method isdir (line 592) | def isdir(self, filepath: Union[str, Path]) -> bool:
    method isfile (line 605) | def isfile(self, filepath: Union[str, Path]) -> bool:
    method join_path (line 617) | def join_path(self, filepath: Union[str, Path],
    method get_local_path (line 633) | def get_local_path(
    method list_dir_or_file (line 638) | def list_dir_or_file(self,
  class HTTPBackend (line 691) | class HTTPBackend(BaseStorageBackend):
    method get (line 694) | def get(self, filepath):
    method get_text (line 698) | def get_text(self, filepath, encoding='utf-8'):
    method get_local_path (line 703) | def get_local_path(self, filepath: str) -> Iterable[str]:
  class FileClient (line 729) | class FileClient:
    method __new__ (line 787) | def __new__(cls, backend=None, prefix=None, **kwargs):
    method name (line 823) | def name(self):
    method allow_symlink (line 827) | def allow_symlink(self):
    method parse_uri_prefix (line 831) | def parse_uri_prefix(uri: Union[str, Path]) -> Optional[str]:
    method infer_client (line 858) | def infer_client(cls,
    method _register_backend (line 886) | def _register_backend(cls, name, backend, force=False, prefixes=None):
    method register_backend (line 922) | def register_backend(cls, name, backend=None, force=False, prefixes=No...
    method get (line 976) | def get(self, filepath: Union[str, Path]) -> Union[bytes, memoryview]:
    method get_text (line 994) | def get_text(self, filepath: Union[str, Path], encoding='utf-8') -> str:
    method put (line 1007) | def put(self, obj: bytes, filepath: Union[str, Path]) -> None:
    method put_text (line 1020) | def put_text(self, obj: str, filepath: Union[str, Path]) -> None:
    method remove (line 1035) | def remove(self, filepath: Union[str, Path]) -> None:
    method exists (line 1043) | def exists(self, filepath: Union[str, Path]) -> bool:
    method isdir (line 1054) | def isdir(self, filepath: Union[str, Path]) -> bool:
    method isfile (line 1067) | def isfile(self, filepath: Union[str, Path]) -> bool:
    method join_path (line 1079) | def join_path(self, filepath: Union[str, Path],
    method get_local_path (line 1095) | def get_local_path(self, filepath: Union[str, Path]) -> Iterable[str]:
    method list_dir_or_file (line 1123) | def list_dir_or_file(self,

FILE: lavis/common/annotator/uniformer/mmcv/fileio/handlers/base.py
  class BaseFileHandler (line 5) | class BaseFileHandler(metaclass=ABCMeta):
    method load_from_fileobj (line 13) | def load_from_fileobj(self, file, **kwargs):
    method dump_to_fileobj (line 17) | def dump_to_fileobj(self, obj, file, **kwargs):
    method dump_to_str (line 21) | def dump_to_str(self, obj, **kwargs):
    method load_from_path (line 24) | def load_from_path(self, filepath, mode='r', **kwargs):
    method dump_to_path (line 28) | def dump_to_path(self, obj, filepath, mode='w', **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/fileio/handlers/json_handler.py
  function set_default (line 9) | def set_default(obj):
  class JsonHandler (line 25) | class JsonHandler(BaseFileHandler):
    method load_from_fileobj (line 27) | def load_from_fileobj(self, file):
    method dump_to_fileobj (line 30) | def dump_to_fileobj(self, obj, file, **kwargs):
    method dump_to_str (line 34) | def dump_to_str(self, obj, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/fileio/handlers/pickle_handler.py
  class PickleHandler (line 7) | class PickleHandler(BaseFileHandler):
    method load_from_fileobj (line 11) | def load_from_fileobj(self, file, **kwargs):
    method load_from_path (line 14) | def load_from_path(self, filepath, **kwargs):
    method dump_to_str (line 18) | def dump_to_str(self, obj, **kwargs):
    method dump_to_fileobj (line 22) | def dump_to_fileobj(self, obj, file, **kwargs):
    method dump_to_path (line 26) | def dump_to_path(self, obj, filepath, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/fileio/handlers/yaml_handler.py
  class YamlHandler (line 12) | class YamlHandler(BaseFileHandler):
    method load_from_fileobj (line 14) | def load_from_fileobj(self, file, **kwargs):
    method dump_to_fileobj (line 18) | def dump_to_fileobj(self, obj, file, **kwargs):
    method dump_to_str (line 22) | def dump_to_str(self, obj, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/fileio/io.py
  function load (line 18) | def load(file, file_format=None, file_client_args=None, **kwargs):
  function dump (line 69) | def dump(obj, file=None, file_format=None, file_client_args=None, **kwar...
  function _register_handler (line 126) | def _register_handler(handler, file_formats):
  function register_handler (line 145) | def register_handler(file_formats, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/fileio/parse.py
  function list_from_file (line 8) | def list_from_file(filename,
  function dict_from_file (line 55) | def dict_from_file(filename,

FILE: lavis/common/annotator/uniformer/mmcv/image/colorspace.py
  function imconvert (line 6) | def imconvert(img, src, dst):
  function bgr2gray (line 22) | def bgr2gray(img, keepdim=False):
  function rgb2gray (line 39) | def rgb2gray(img, keepdim=False):
  function gray2bgr (line 56) | def gray2bgr(img):
  function gray2rgb (line 70) | def gray2rgb(img):
  function _convert_input_type_range (line 84) | def _convert_input_type_range(img):
  function _convert_output_type_range (line 112) | def _convert_output_type_range(img, dst_type):
  function rgb2ycbcr (line 143) | def rgb2ycbcr(img, y_only=False):
  function bgr2ycbcr (line 177) | def bgr2ycbcr(img, y_only=False):
  function ycbcr2rgb (line 211) | def ycbcr2rgb(img):
  function ycbcr2bgr (line 243) | def ycbcr2bgr(img):
  function convert_color_factory (line 275) | def convert_color_factory(src, dst):

FILE: lavis/common/annotator/uniformer/mmcv/image/geometric.py
  function _scale_size (line 16) | def _scale_size(size, scale):
  function imresize (line 51) | def imresize(img,
  function imresize_to_multiple (line 98) | def imresize_to_multiple(img,
  function imresize_like (line 162) | def imresize_like(img,
  function rescale_size (line 184) | def rescale_size(old_size, scale, return_scale=False):
  function imrescale (line 221) | def imrescale(img,
  function imflip (line 252) | def imflip(img, direction='horizontal'):
  function imflip_ (line 272) | def imflip_(img, direction='horizontal'):
  function imrotate (line 292) | def imrotate(img,
  function bbox_clip (line 342) | def bbox_clip(bboxes, img_shape):
  function bbox_scaling (line 360) | def bbox_scaling(bboxes, scale, clip_shape=None):
  function imcrop (line 386) | def imcrop(img, bboxes, scale=1.0, pad_fill=None):
  function impad (line 440) | def impad(img,
  function impad_to_multiple (line 522) | def impad_to_multiple(img, divisor, pad_val=0):
  function cutout (line 538) | def cutout(img, shape, pad_val=0):
  function _get_shear_matrix (line 593) | def _get_shear_matrix(magnitude, direction='horizontal'):
  function imshear (line 611) | def imshear(img,
  function _get_translate_matrix (line 662) | def _get_translate_matrix(offset, direction='horizontal'):
  function imtranslate (line 680) | def imtranslate(img,

FILE: lavis/common/annotator/uniformer/mmcv/image/io.py
  function use_backend (line 43) | def use_backend(backend):
  function _jpegflag (line 69) | def _jpegflag(flag='color', channel_order='bgr'):
  function _pillow2array (line 85) | def _pillow2array(img, flag='color', channel_order='bgr'):
  function imread (line 140) | def imread(img_or_path, flag='color', channel_order='bgr', backend=None):
  function imfrombytes (line 203) | def imfrombytes(content, flag='color', channel_order='bgr', backend=None):
  function imwrite (line 242) | def imwrite(img, file_path, params=None, auto_mkdir=True):

FILE: lavis/common/annotator/uniformer/mmcv/image/misc.py
  function tensor2imgs (line 12) | def tensor2imgs(tensor, mean=(0, 0, 0), std=(1, 1, 1), to_rgb=True):

FILE: lavis/common/annotator/uniformer/mmcv/image/photometric.py
  function imnormalize (line 9) | def imnormalize(img, mean, std, to_rgb=True):
  function imnormalize_ (line 25) | def imnormalize_(img, mean, std, to_rgb=True):
  function imdenormalize (line 48) | def imdenormalize(img, mean, std, to_bgr=True):
  function iminvert (line 59) | def iminvert(img):
  function solarize (line 71) | def solarize(img, thr=128):
  function posterize (line 85) | def posterize(img, bits):
  function adjust_color (line 100) | def adjust_color(img, alpha=1, beta=None, gamma=0):
  function imequalize (line 131) | def imequalize(img):
  function adjust_brightness (line 176) | def adjust_brightness(img, factor=1.):
  function adjust_contrast (line 208) | def adjust_contrast(img, factor=1.):
  function auto_contrast (line 238) | def auto_contrast(img, cutoff=0):
  function adjust_sharpness (line 294) | def adjust_sharpness(img, factor=1., kernel=None):
  function adjust_lighting (line 338) | def adjust_lighting(img, eigval, eigvec, alphastd=0.1, to_rgb=True):
  function lut_transform (line 381) | def lut_transform(img, lut_table):
  function clahe (line 405) | def clahe(img, clip_limit=40.0, tile_grid_size=(8, 8)):

FILE: lavis/common/annotator/uniformer/mmcv/ops/assign_score_withk.py
  class AssignScoreWithK (line 9) | class AssignScoreWithK(Function):
    method forward (line 29) | def forward(ctx,
    method backward (line 81) | def backward(ctx, grad_out):

FILE: lavis/common/annotator/uniformer/mmcv/ops/ball_query.py
  class BallQuery (line 10) | class BallQuery(Function):
    method forward (line 14) | def forward(ctx, min_radius: float, max_radius: float, sample_num: int,
    method backward (line 51) | def backward(ctx, a=None):

FILE: lavis/common/annotator/uniformer/mmcv/ops/bbox.py
  function bbox_overlaps (line 7) | def bbox_overlaps(bboxes1, bboxes2, mode='iou', aligned=False, offset=0):

FILE: lavis/common/annotator/uniformer/mmcv/ops/border_align.py
  class BorderAlignFunction (line 16) | class BorderAlignFunction(Function):
    method symbolic (line 19) | def symbolic(g, input, boxes, pool_size):
    method forward (line 24) | def forward(ctx, input, boxes, pool_size):
    method backward (line 48) | def backward(ctx, grad_output):
  class BorderAlign (line 65) | class BorderAlign(nn.Module):
    method __init__ (line 88) | def __init__(self, pool_size):
    method forward (line 92) | def forward(self, input, boxes):
    method __repr__ (line 106) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmcv/ops/box_iou_rotated.py
  function box_iou_rotated (line 7) | def box_iou_rotated(bboxes1, bboxes2, mode='iou', aligned=False):

FILE: lavis/common/annotator/uniformer/mmcv/ops/carafe.py
  class CARAFENaiveFunction (line 17) | class CARAFENaiveFunction(Function):
    method symbolic (line 20) | def symbolic(g, features, masks, kernel_size, group_size, scale_factor):
    method forward (line 30) | def forward(ctx, features, masks, kernel_size, group_size, scale_factor):
    method backward (line 58) | def backward(ctx, grad_output):
  class CARAFENaive (line 84) | class CARAFENaive(Module):
    method __init__ (line 86) | def __init__(self, kernel_size, group_size, scale_factor):
    method forward (line 95) | def forward(self, features, masks):
  class CARAFEFunction (line 100) | class CARAFEFunction(Function):
    method symbolic (line 103) | def symbolic(g, features, masks, kernel_size, group_size, scale_factor):
    method forward (line 113) | def forward(ctx, features, masks, kernel_size, group_size, scale_factor):
    method backward (line 147) | def backward(ctx, grad_output):
  class CARAFE (line 180) | class CARAFE(Module):
    method __init__ (line 194) | def __init__(self, kernel_size, group_size, scale_factor):
    method forward (line 203) | def forward(self, features, masks):
  class CARAFEPack (line 209) | class CARAFEPack(nn.Module):
    method __init__ (line 230) | def __init__(self,
    method init_weights (line 258) | def init_weights(self):
    method kernel_normalizer (line 264) | def kernel_normalizer(self, mask):
    method feature_reassemble (line 277) | def feature_reassemble(self, x, mask):
    method forward (line 281) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/ops/cc_attention.py
  function NEG_INF_DIAG (line 9) | def NEG_INF_DIAG(n, device):
  class CrissCrossAttention (line 19) | class CrissCrossAttention(nn.Module):
    method __init__ (line 44) | def __init__(self, in_channels):
    method forward (line 52) | def forward(self, x):
    method __repr__ (line 80) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmcv/ops/contour_expand.py
  function contour_expand (line 10) | def contour_expand(kernel_mask, internal_kernel_label, min_kernel_area,

FILE: lavis/common/annotator/uniformer/mmcv/ops/corner_pool.py
  class TopPoolFunction (line 17) | class TopPoolFunction(Function):
    method symbolic (line 20) | def symbolic(g, input):
    method forward (line 26) | def forward(ctx, input):
    method backward (line 32) | def backward(ctx, grad_output):
  class BottomPoolFunction (line 38) | class BottomPoolFunction(Function):
    method symbolic (line 41) | def symbolic(g, input):
    method forward (line 47) | def forward(ctx, input):
    method backward (line 53) | def backward(ctx, grad_output):
  class LeftPoolFunction (line 59) | class LeftPoolFunction(Function):
    method symbolic (line 62) | def symbolic(g, input):
    method forward (line 68) | def forward(ctx, input):
    method backward (line 74) | def backward(ctx, grad_output):
  class RightPoolFunction (line 80) | class RightPoolFunction(Function):
    method symbolic (line 83) | def symbolic(g, input):
    method forward (line 89) | def forward(ctx, input):
    method backward (line 95) | def backward(ctx, grad_output):
  class CornerPool (line 101) | class CornerPool(nn.Module):
    method __init__ (line 136) | def __init__(self, mode):
    method forward (line 142) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/ops/correlation.py
  class CorrelationFunction (line 14) | class CorrelationFunction(Function):
    method forward (line 17) | def forward(ctx,
    method backward (line 63) | def backward(ctx, grad_output):
    method _output_size (line 96) | def _output_size(ctx, input1):
  class Correlation (line 114) | class Correlation(nn.Module):
    method __init__ (line 167) | def __init__(self,
    method forward (line 182) | def forward(self, input1: Tensor, input2: Tensor) -> Tensor:
    method __repr__ (line 188) | def __repr__(self) -> str:

FILE: lavis/common/annotator/uniformer/mmcv/ops/deform_conv.py
  class DeformConv2dFunction (line 22) | class DeformConv2dFunction(Function):
    method symbolic (line 25) | def symbolic(g,
    method forward (line 50) | def forward(ctx,
    method backward (line 114) | def backward(ctx, grad_output):
    method _output_size (line 173) | def _output_size(ctx, input, weight):
  class DeformConv2d (line 192) | class DeformConv2d(nn.Module):
    method __init__ (line 228) | def __init__(self,
    method reset_parameters (line 269) | def reset_parameters(self):
    method forward (line 276) | def forward(self, x: Tensor, offset: Tensor) -> Tensor:
    method __repr__ (line 315) | def __repr__(self):
  class DeformConv2dPack (line 331) | class DeformConv2dPack(DeformConv2d):
    method __init__ (line 358) | def __init__(self, *args, **kwargs):
    method init_offset (line 370) | def init_offset(self):
    method forward (line 374) | def forward(self, x):
    method _load_from_state_dict (line 380) | def _load_from_state_dict(self, state_dict, prefix, local_metadata, st...

FILE: lavis/common/annotator/uniformer/mmcv/ops/deform_roi_pool.py
  class DeformRoIPoolFunction (line 13) | class DeformRoIPoolFunction(Function):
    method symbolic (line 16) | def symbolic(g, input, rois, offset, output_size, spatial_scale,
    method forward (line 30) | def forward(ctx,
    method backward (line 67) | def backward(ctx, grad_output):
  class DeformRoIPool (line 92) | class DeformRoIPool(nn.Module):
    method __init__ (line 94) | def __init__(self,
    method forward (line 105) | def forward(self, input, rois, offset=None):
  class DeformRoIPoolPack (line 111) | class DeformRoIPoolPack(DeformRoIPool):
    method __init__ (line 113) | def __init__(self,
    method forward (line 138) | def forward(self, input, rois):
  class ModulatedDeformRoIPoolPack (line 152) | class ModulatedDeformRoIPoolPack(DeformRoIPool):
    method __init__ (line 154) | def __init__(self,
    method forward (line 190) | def forward(self, input, rois):

FILE: lavis/common/annotator/uniformer/mmcv/ops/deprecated_wrappers.py
  class Conv2d_deprecated (line 9) | class Conv2d_deprecated(Conv2d):
    method __init__ (line 11) | def __init__(self, *args, **kwargs):
  class ConvTranspose2d_deprecated (line 18) | class ConvTranspose2d_deprecated(ConvTranspose2d):
    method __init__ (line 20) | def __init__(self, *args, **kwargs):
  class MaxPool2d_deprecated (line 28) | class MaxPool2d_deprecated(MaxPool2d):
    method __init__ (line 30) | def __init__(self, *args, **kwargs):
  class Linear_deprecated (line 37) | class Linear_deprecated(Linear):
    method __init__ (line 39) | def __init__(self, *args, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/ops/focal_loss.py
  class SigmoidFocalLossFunction (line 15) | class SigmoidFocalLossFunction(Function):
    method symbolic (line 18) | def symbolic(g, input, target, gamma, alpha, weight, reduction):
    method forward (line 29) | def forward(ctx,
    method backward (line 66) | def backward(ctx, grad_output):
  class SigmoidFocalLoss (line 88) | class SigmoidFocalLoss(nn.Module):
    method __init__ (line 90) | def __init__(self, gamma, alpha, weight=None, reduction='mean'):
    method forward (line 97) | def forward(self, input, target):
    method __repr__ (line 101) | def __repr__(self):
  class SoftmaxFocalLossFunction (line 109) | class SoftmaxFocalLossFunction(Function):
    method symbolic (line 112) | def symbolic(g, input, target, gamma, alpha, weight, reduction):
    method forward (line 123) | def forward(ctx,
    method backward (line 171) | def backward(ctx, grad_output):
  class SoftmaxFocalLoss (line 194) | class SoftmaxFocalLoss(nn.Module):
    method __init__ (line 196) | def __init__(self, gamma, alpha, weight=None, reduction='mean'):
    method forward (line 203) | def forward(self, input, target):
    method __repr__ (line 207) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmcv/ops/furthest_point_sample.py
  class FurthestPointSampling (line 12) | class FurthestPointSampling(Function):
    method forward (line 17) | def forward(ctx, points_xyz: torch.Tensor,
    method backward (line 46) | def backward(xyz, a=None):
  class FurthestPointSamplingWithDist (line 50) | class FurthestPointSamplingWithDist(Function):
    method forward (line 55) | def forward(ctx, points_dist: torch.Tensor,
    method backward (line 78) | def backward(xyz, a=None):

FILE: lavis/common/annotator/uniformer/mmcv/ops/fused_bias_leakyrelu.py
  class FusedBiasLeakyReLUFunctionBackward (line 108) | class FusedBiasLeakyReLUFunctionBackward(Function):
    method forward (line 116) | def forward(ctx, grad_output, out, negative_slope, scale):
    method backward (line 142) | def backward(ctx, gradgrad_input, gradgrad_bias):
  class FusedBiasLeakyReLUFunction (line 160) | class FusedBiasLeakyReLUFunction(Function):
    method forward (line 163) | def forward(ctx, input, bias, negative_slope, scale):
    method backward (line 181) | def backward(ctx, grad_output):
  class FusedBiasLeakyReLU (line 190) | class FusedBiasLeakyReLU(nn.Module):
    method __init__ (line 213) | def __init__(self, num_channels, negative_slope=0.2, scale=2**0.5):
    method forward (line 220) | def forward(self, input):
  function fused_bias_leakyrelu (line 225) | def fused_bias_leakyrelu(input, bias, negative_slope=0.2, scale=2**0.5):
  function bias_leakyrelu_ref (line 257) | def bias_leakyrelu_ref(x, bias, negative_slope=0.2, scale=2**0.5):

FILE: lavis/common/annotator/uniformer/mmcv/ops/gather_points.py
  class GatherPoints (line 10) | class GatherPoints(Function):
    method forward (line 14) | def forward(ctx, features: torch.Tensor,
    method backward (line 40) | def backward(ctx, grad_out):

FILE: lavis/common/annotator/uniformer/mmcv/ops/group_points.py
  class QueryAndGroup (line 16) | class QueryAndGroup(nn.Module):
    method __init__ (line 39) | def __init__(self,
    method forward (line 67) | def forward(self, points_xyz, center_xyz, features=None):
  class GroupAll (line 134) | class GroupAll(nn.Module):
    method __init__ (line 141) | def __init__(self, use_xyz: bool = True):
    method forward (line 145) | def forward(self,
  class GroupingOperation (line 173) | class GroupingOperation(Function):
    method forward (line 177) | def forward(ctx, features: torch.Tensor,
    method backward (line 202) | def backward(ctx,

FILE: lavis/common/annotator/uniformer/mmcv/ops/info.py
  function get_compiler_version (line 10) | def get_compiler_version():
  function get_compiling_cuda_version (line 13) | def get_compiling_cuda_version():
  function get_compiler_version (line 20) | def get_compiler_version():
  function get_compiling_cuda_version (line 23) | def get_compiling_cuda_version():
  function get_onnxruntime_op_path (line 27) | def get_onnxruntime_op_path():

FILE: lavis/common/annotator/uniformer/mmcv/ops/iou3d.py
  function boxes_iou_bev (line 12) | def boxes_iou_bev(boxes_a, boxes_b):
  function nms_bev (line 31) | def nms_bev(boxes, scores, thresh, pre_max_size=None, post_max_size=None):
  function nms_normal_bev (line 65) | def nms_normal_bev(boxes, scores, thresh):

FILE: lavis/common/annotator/uniformer/mmcv/ops/knn.py
  class KNN (line 9) | class KNN(Function):
    method forward (line 18) | def forward(ctx,
    method backward (line 73) | def backward(ctx, a=None):

FILE: lavis/common/annotator/uniformer/mmcv/ops/masked_conv.py
  class MaskedConv2dFunction (line 16) | class MaskedConv2dFunction(Function):
    method symbolic (line 19) | def symbolic(g, features, mask, weight, bias, padding, stride):
    method forward (line 30) | def forward(ctx, features, mask, weight, bias, padding=0, stride=1):
    method backward (line 79) | def backward(ctx, grad_output):
  class MaskedConv2d (line 86) | class MaskedConv2d(nn.Conv2d):
    method __init__ (line 93) | def __init__(self,
    method forward (line 106) | def forward(self, input, mask=None):

FILE: lavis/common/annotator/uniformer/mmcv/ops/merge_cells.py
  class BaseMergeCell (line 11) | class BaseMergeCell(nn.Module):
    method __init__ (line 43) | def __init__(self,
    method _build_input_conv (line 78) | def _build_input_conv(self, channel, conv_cfg, norm_cfg):
    method _binary_op (line 89) | def _binary_op(self, x1, x2):
    method _resize (line 92) | def _resize(self, x, size):
    method forward (line 103) | def forward(self, x1, x2, out_size=None):
  class SumCell (line 121) | class SumCell(BaseMergeCell):
    method __init__ (line 123) | def __init__(self, in_channels, out_channels, **kwargs):
    method _binary_op (line 126) | def _binary_op(self, x1, x2):
  class ConcatCell (line 130) | class ConcatCell(BaseMergeCell):
    method __init__ (line 132) | def __init__(self, in_channels, out_channels, **kwargs):
    method _binary_op (line 136) | def _binary_op(self, x1, x2):
  class GlobalPoolingCell (line 141) | class GlobalPoolingCell(BaseMergeCell):
    method __init__ (line 143) | def __init__(self, in_channels=None, out_channels=None, **kwargs):
    method _binary_op (line 147) | def _binary_op(self, x1, x2):

FILE: lavis/common/annotator/uniformer/mmcv/ops/modulated_deform_conv.py
  class ModulatedDeformConv2dFunction (line 19) | class ModulatedDeformConv2dFunction(Function):
    method symbolic (line 22) | def symbolic(g, input, offset, mask, weight, bias, stride, padding,
    method forward (line 37) | def forward(ctx,
    method backward (line 97) | def backward(ctx, grad_output):
    method _output_size (line 137) | def _output_size(ctx, input, weight):
  class ModulatedDeformConv2d (line 156) | class ModulatedDeformConv2d(nn.Module):
    method __init__ (line 160) | def __init__(self,
    method init_weights (line 192) | def init_weights(self):
    method forward (line 201) | def forward(self, x, offset, mask):
  class ModulatedDeformConv2dPack (line 209) | class ModulatedDeformConv2dPack(ModulatedDeformConv2d):
    method __init__ (line 228) | def __init__(self, *args, **kwargs):
    method init_weights (line 240) | def init_weights(self):
    method forward (line 246) | def forward(self, x):
    method _load_from_state_dict (line 256) | def _load_from_state_dict(self, state_dict, prefix, local_metadata, st...

FILE: lavis/common/annotator/uniformer/mmcv/ops/multi_scale_deform_attn.py
  class MultiScaleDeformableAttnFunction (line 20) | class MultiScaleDeformableAttnFunction(Function):
    method forward (line 23) | def forward(ctx, value, value_spatial_shapes, value_level_start_index,
    method backward (line 61) | def backward(ctx, grad_output):
  function multi_scale_deformable_attn_pytorch (line 94) | def multi_scale_deformable_attn_pytorch(value, value_spatial_shapes,
  class MultiScaleDeformableAttention (line 155) | class MultiScaleDeformableAttention(BaseModule):
    method __init__ (line 182) | def __init__(self,
    method init_weights (line 230) | def init_weights(self):
    method forward (line 252) | def forward(self,

FILE: lavis/common/annotator/uniformer/mmcv/ops/nms.py
  class NMSop (line 14) | class NMSop(torch.autograd.Function):
    method forward (line 17) | def forward(ctx, bboxes, scores, iou_threshold, offset, score_threshold,
    method symbolic (line 36) | def symbolic(g, bboxes, scores, iou_threshold, offset, score_threshold,
  class SoftNMSop (line 82) | class SoftNMSop(torch.autograd.Function):
    method forward (line 85) | def forward(ctx, boxes, scores, iou_threshold, sigma, min_score, method,
    method symbolic (line 100) | def symbolic(g, boxes, scores, iou_threshold, sigma, min_score, method,
  function nms (line 118) | def nms(boxes, scores, iou_threshold, offset=0, score_threshold=0, max_n...
  function soft_nms (line 181) | def soft_nms(boxes,
  function batched_nms (line 260) | def batched_nms(boxes, scores, idxs, nms_cfg, class_agnostic=False):
  function nms_match (line 339) | def nms_match(dets, iou_threshold):
  function nms_rotated (line 376) | def nms_rotated(dets, scores, iou_threshold, labels=None):

FILE: lavis/common/annotator/uniformer/mmcv/ops/pixel_group.py
  function pixel_group (line 10) | def pixel_group(score, mask, embedding, kernel_label, kernel_contour,

FILE: lavis/common/annotator/uniformer/mmcv/ops/point_sample.py
  function bilinear_grid_sample (line 12) | def bilinear_grid_sample(im, grid, align_corners=False):
  function is_in_onnx_export_without_custom_ops (line 87) | def is_in_onnx_export_without_custom_ops():
  function normalize (line 94) | def normalize(grid):
  function denormalize (line 105) | def denormalize(grid):
  function generate_grid (line 116) | def generate_grid(num_grid, size, device):
  function rel_roi_point_to_abs_img_point (line 137) | def rel_roi_point_to_abs_img_point(rois, rel_roi_points):
  function get_shape_from_feature_map (line 168) | def get_shape_from_feature_map(x):
  function abs_img_point_to_rel_img_point (line 186) | def abs_img_point_to_rel_img_point(abs_img_points, img, spatial_scale=1.):
  function rel_roi_point_to_rel_img_point (line 216) | def rel_roi_point_to_rel_img_point(rois,
  function point_sample (line 242) | def point_sample(input, points, align_corners=False, **kwargs):
  class SimpleRoIAlign (line 276) | class SimpleRoIAlign(nn.Module):
    method __init__ (line 278) | def __init__(self, output_size, spatial_scale, aligned=True):
    method forward (line 296) | def forward(self, features, rois):
    method __repr__ (line 332) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmcv/ops/points_in_boxes.py
  function points_in_boxes_part (line 11) | def points_in_boxes_part(points, boxes):
  function points_in_boxes_cpu (line 58) | def points_in_boxes_cpu(points, boxes):
  function points_in_boxes_all (line 95) | def points_in_boxes_all(points, boxes):

FILE: lavis/common/annotator/uniformer/mmcv/ops/points_sampler.py
  function calc_square_dist (line 11) | def calc_square_dist(point_feat_a, point_feat_b, norm=True):
  function get_sampler_cls (line 37) | def get_sampler_cls(sampler_type):
  class PointsSampler (line 60) | class PointsSampler(nn.Module):
    method __init__ (line 74) | def __init__(self,
    method forward (line 92) | def forward(self, points_xyz, features):
  class DFPSSampler (line 133) | class DFPSSampler(nn.Module):
    method __init__ (line 136) | def __init__(self):
    method forward (line 139) | def forward(self, points, features, npoint):
  class FFPSSampler (line 145) | class FFPSSampler(nn.Module):
    method __init__ (line 148) | def __init__(self):
    method forward (line 151) | def forward(self, points, features, npoint):
  class FSSampler (line 162) | class FSSampler(nn.Module):
    method __init__ (line 165) | def __init__(self):
    method forward (line 168) | def forward(self, points, features, npoint):

FILE: lavis/common/annotator/uniformer/mmcv/ops/psa_mask.py
  class PSAMaskFunction (line 12) | class PSAMaskFunction(Function):
    method symbolic (line 15) | def symbolic(g, input, psa_type, mask_size):
    method forward (line 23) | def forward(ctx, input, psa_type, mask_size):
    method backward (line 48) | def backward(ctx, grad_output):
  class PSAMask (line 72) | class PSAMask(nn.Module):
    method __init__ (line 74) | def __init__(self, psa_type, mask_size=None):
    method forward (line 85) | def forward(self, input):
    method __repr__ (line 88) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmcv/ops/roi_align.py
  class RoIAlignFunction (line 14) | class RoIAlignFunction(Function):
    method symbolic (line 17) | def symbolic(g, input, rois, output_size, spatial_scale, sampling_ratio,
    method forward (line 64) | def forward(ctx,
    method backward (line 110) | def backward(ctx, grad_output):
  class RoIAlign (line 133) | class RoIAlign(nn.Module):
    method __init__ (line 176) | def __init__(self,
    method forward (line 192) | def forward(self, input, rois):
    method __repr__ (line 215) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmcv/ops/roi_align_rotated.py
  class RoIAlignRotatedFunction (line 11) | class RoIAlignRotatedFunction(Function):
    method symbolic (line 14) | def symbolic(g, features, rois, out_size, spatial_scale, sample_num,
    method forward (line 39) | def forward(ctx,
    method backward (line 82) | def backward(ctx, grad_output):
  class RoIAlignRotated (line 116) | class RoIAlignRotated(nn.Module):
    method __init__ (line 159) | def __init__(self,
    method forward (line 173) | def forward(self, features, rois):

FILE: lavis/common/annotator/uniformer/mmcv/ops/roi_pool.py
  class RoIPoolFunction (line 14) | class RoIPoolFunction(Function):
    method symbolic (line 17) | def symbolic(g, input, rois, output_size, spatial_scale):
    method forward (line 26) | def forward(ctx, input, rois, output_size, spatial_scale=1.0):
    method backward (line 52) | def backward(ctx, grad_output):
  class RoIPool (line 71) | class RoIPool(nn.Module):
    method __init__ (line 73) | def __init__(self, output_size, spatial_scale=1.0):
    method forward (line 79) | def forward(self, input, rois):
    method __repr__ (line 82) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmcv/ops/roiaware_pool3d.py
  class RoIAwarePool3d (line 13) | class RoIAwarePool3d(nn.Module):
    method __init__ (line 28) | def __init__(self, out_size, max_pts_per_voxel=128, mode='max'):
    method forward (line 37) | def forward(self, rois, pts, pts_feature):
  class RoIAwarePool3dFunction (line 54) | class RoIAwarePool3dFunction(Function):
    method forward (line 57) | def forward(ctx, rois, pts, pts_feature, out_size, max_pts_per_voxel,
    method backward (line 105) | def backward(ctx, grad_out):

FILE: lavis/common/annotator/uniformer/mmcv/ops/roipoint_pool3d.py
  class RoIPointPool3d (line 9) | class RoIPointPool3d(nn.Module):
    method __init__ (line 20) | def __init__(self, num_sampled_points=512):
    method forward (line 24) | def forward(self, points, point_features, boxes3d):
  class RoIPointPool3dFunction (line 41) | class RoIPointPool3dFunction(Function):
    method forward (line 44) | def forward(ctx, points, point_features, boxes3d, num_sampled_points=5...
    method backward (line 76) | def backward(ctx, grad_out):

FILE: lavis/common/annotator/uniformer/mmcv/ops/saconv.py
  class SAConv2d (line 12) | class SAConv2d(ConvAWS2d):
    method __init__ (line 37) | def __init__(self,
    method init_weights (line 81) | def init_weights(self):
    method forward (line 90) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmcv/ops/scatter_points.py
  class _DynamicScatter (line 13) | class _DynamicScatter(Function):
    method forward (line 16) | def forward(ctx, feats, coors, reduce_type='max'):
    method backward (line 44) | def backward(ctx, grad_voxel_feats, grad_voxel_coors=None):
  class DynamicScatter (line 59) | class DynamicScatter(nn.Module):
    method __init__ (line 75) | def __init__(self, voxel_size, point_cloud_range, average_points: bool):
    method forward_single (line 82) | def forward_single(self, points, coors):
    method forward (line 98) | def forward(self, points, coors):
    method __repr__ (line 129) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmcv/ops/sync_bn.py
  class SyncBatchNormFunction (line 19) | class SyncBatchNormFunction(Function):
    method symbolic (line 22) | def symbolic(g, input, running_mean, running_var, weight, bias, momentum,
    method forward (line 38) | def forward(self, input, running_mean, running_var, weight, bias, mome...
    method backward (line 129) | def backward(self, grad_output):
  class SyncBatchNorm (line 159) | class SyncBatchNorm(Module):
    method __init__ (line 193) | def __init__(self,
    method reset_running_stats (line 230) | def reset_running_stats(self):
    method reset_parameters (line 236) | def reset_parameters(self):
    method forward (line 242) | def forward(self, input):
    method __repr__ (line 270) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmcv/ops/three_interpolate.py
  class ThreeInterpolate (line 12) | class ThreeInterpolate(Function):
    method forward (line 20) | def forward(ctx, features: torch.Tensor, indices: torch.Tensor,
    method backward (line 47) | def backward(

FILE: lavis/common/annotator/uniformer/mmcv/ops/three_nn.py
  class ThreeNN (line 11) | class ThreeNN(Function):
    method forward (line 19) | def forward(ctx, target: torch.Tensor,
    method backward (line 47) | def backward(ctx, a=None, b=None):

FILE: lavis/common/annotator/uniformer/mmcv/ops/tin_shift.py
  class TINShiftFunction (line 17) | class TINShiftFunction(Function):
    method forward (line 20) | def forward(ctx, input, shift):
    method backward (line 35) | def backward(ctx, grad_output):
  class TINShift (line 48) | class TINShift(nn.Module):
    method forward (line 58) | def forward(self, input, shift):

FILE: lavis/common/annotator/uniformer/mmcv/ops/upfirdn2d.py
  class UpFirDn2dBackward (line 108) | class UpFirDn2dBackward(Function):
    method forward (line 111) | def forward(ctx, grad_output, kernel, grad_kernel, up, down, pad, g_pad,
    method backward (line 152) | def backward(ctx, gradgrad_input):
  class UpFirDn2d (line 177) | class UpFirDn2d(Function):
    method forward (line 180) | def forward(ctx, input, kernel, up, down, pad):
    method backward (line 225) | def backward(ctx, grad_output):
  function upfirdn2d (line 243) | def upfirdn2d(input, kernel, up=1, down=1, pad=(0, 0)):
  function upfirdn2d_native (line 290) | def upfirdn2d_native(input, kernel, up_x, up_y, down_x, down_y, pad_x0, ...

FILE: lavis/common/annotator/uniformer/mmcv/ops/voxelize.py
  class _Voxelization (line 13) | class _Voxelization(Function):
    method forward (line 16) | def forward(ctx,
  class Voxelization (line 71) | class Voxelization(nn.Module):
    method __init__ (line 89) | def __init__(self,
    method forward (line 116) | def forward(self, input):
    method __repr__ (line 125) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmcv/parallel/_functions.py
  function scatter (line 6) | def scatter(input, devices, streams=None):
  function synchronize_stream (line 34) | def synchronize_stream(output, devices, streams):
  function get_input_device (line 51) | def get_input_device(input):
  class Scatter (line 64) | class Scatter:
    method forward (line 67) | def forward(target_gpus, input):

FILE: lavis/common/annotator/uniformer/mmcv/parallel/collate.py
  function collate (line 11) | def collate(batch, samples_per_gpu=1):

FILE: lavis/common/annotator/uniformer/mmcv/parallel/data_container.py
  function assert_tensor_type (line 7) | def assert_tensor_type(func):
  class DataContainer (line 20) | class DataContainer:
    method __init__ (line 37) | def __init__(self,
    method __repr__ (line 50) | def __repr__(self):
    method __len__ (line 53) | def __len__(self):
    method data (line 57) | def data(self):
    method datatype (line 61) | def datatype(self):
    method cpu_only (line 68) | def cpu_only(self):
    method stack (line 72) | def stack(self):
    method padding_value (line 76) | def padding_value(self):
    method pad_dims (line 80) | def pad_dims(self):
    method size (line 84) | def size(self, *args, **kwargs):
    method dim (line 88) | def dim(self):

FILE: lavis/common/annotator/uniformer/mmcv/parallel/data_parallel.py
  class MMDataParallel (line 9) | class MMDataParallel(DataParallel):
    method __init__ (line 26) | def __init__(self, *args, dim=0, **kwargs):
    method forward (line 30) | def forward(self, *inputs, **kwargs):
    method scatter (line 44) | def scatter(self, inputs, kwargs, device_ids):
    method train_step (line 47) | def train_step(self, *inputs, **kwargs):
    method val_step (line 69) | def val_step(self, *inputs, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/parallel/distributed.py
  class MMDistributedDataParallel (line 11) | class MMDistributedDataParallel(DistributedDataParallel):
    method to_kwargs (line 21) | def to_kwargs(self, inputs, kwargs, device_id):
    method scatter (line 26) | def scatter(self, inputs, kwargs, device_ids):
    method train_step (line 29) | def train_step(self, *inputs, **kwargs):
    method val_step (line 72) | def val_step(self, *inputs, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/parallel/distributed_deprecated.py
  class MMDistributedDataParallel (line 14) | class MMDistributedDataParallel(nn.Module):
    method __init__ (line 16) | def __init__(self,
    method _dist_broadcast_coalesced (line 29) | def _dist_broadcast_coalesced(self, tensors, buffer_size):
    method _sync_params (line 37) | def _sync_params(self):
    method scatter (line 52) | def scatter(self, inputs, kwargs, device_ids):
    method forward (line 55) | def forward(self, *inputs, **kwargs):
    method train_step (line 60) | def train_step(self, *inputs, **kwargs):
    method val_step (line 66) | def val_step(self, *inputs, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/parallel/scatter_gather.py
  function scatter (line 9) | def scatter(inputs, target_gpus, dim=0):
  function scatter_kwargs (line 49) | def scatter_kwargs(inputs, kwargs, target_gpus, dim=0):

FILE: lavis/common/annotator/uniformer/mmcv/parallel/utils.py
  function is_module_wrapper (line 5) | def is_module_wrapper(module):

FILE: lavis/common/annotator/uniformer/mmcv/runner/base_module.py
  class BaseModule (line 14) | class BaseModule(nn.Module, metaclass=ABCMeta):
    method __init__ (line 33) | def __init__(self, init_cfg=None):
    method is_init (line 53) | def is_init(self):
    method init_weights (line 56) | def init_weights(self):
    method _dump_init_info (line 137) | def _dump_init_info(self, logger_name):
    method __repr__ (line 166) | def __repr__(self):
  class Sequential (line 173) | class Sequential(BaseModule, nn.Sequential):
    method __init__ (line 180) | def __init__(self, *args, init_cfg=None):
  class ModuleList (line 185) | class ModuleList(BaseModule, nn.ModuleList):
    method __init__ (line 193) | def __init__(self, modules=None, init_cfg=None):

FILE: lavis/common/annotator/uniformer/mmcv/runner/base_runner.py
  class BaseRunner (line 21) | class BaseRunner(metaclass=ABCMeta):
    method __init__ (line 51) | def __init__(self,
    method model_name (line 139) | def model_name(self):
    method rank (line 144) | def rank(self):
    method world_size (line 149) | def world_size(self):
    method hooks (line 155) | def hooks(self):
    method epoch (line 160) | def epoch(self):
    method iter (line 165) | def iter(self):
    method inner_iter (line 170) | def inner_iter(self):
    method max_epochs (line 175) | def max_epochs(self):
    method max_iters (line 180) | def max_iters(self):
    method train (line 185) | def train(self):
    method val (line 189) | def val(self):
    method run (line 193) | def run(self, data_loaders, workflow, **kwargs):
    method save_checkpoint (line 197) | def save_checkpoint(self,
    method current_lr (line 205) | def current_lr(self):
    method current_momentum (line 224) | def current_momentum(self):
    method register_hook (line 255) | def register_hook(self, hook, priority='NORMAL'):
    method register_hook_from_cfg (line 283) | def register_hook_from_cfg(self, hook_cfg):
    method call_hook (line 299) | def call_hook(self, fn_name):
    method get_hook_info (line 309) | def get_hook_info(self):
    method load_checkpoint (line 332) | def load_checkpoint(self,
    method resume (line 345) | def resume(self,
    method register_lr_hook (line 399) | def register_lr_hook(self, lr_config):
    method register_momentum_hook (line 420) | def register_momentum_hook(self, momentum_config):
    method register_optimizer_hook (line 441) | def register_optimizer_hook(self, optimizer_config):
    method register_checkpoint_hook (line 451) | def register_checkpoint_hook(self, checkpoint_config):
    method register_logger_hooks (line 461) | def register_logger_hooks(self, log_config):
    method register_timer_hook (line 470) | def register_timer_hook(self, timer_config):
    method register_custom_hooks (line 480) | def register_custom_hooks(self, custom_config):
    method register_profiler_hook (line 493) | def register_profiler_hook(self, profiler_config):
    method register_training_hooks (line 503) | def register_training_hooks(self,

FILE: lavis/common/annotator/uniformer/mmcv/runner/builder.py
  function build_runner_constructor (line 10) | def build_runner_constructor(cfg):
  function build_runner (line 14) | def build_runner(cfg, default_args=None):

FILE: lavis/common/annotator/uniformer/mmcv/runner/checkpoint.py
  function _get_mmcv_home (line 30) | def _get_mmcv_home():
  function load_state_dict (line 41) | def load_state_dict(module, state_dict, strict=False, logger=None):
  function get_torchvision_models (line 109) | def get_torchvision_models():
  function get_external_models (line 121) | def get_external_models():
  function get_mmcls_models (line 135) | def get_mmcls_models():
  function get_deprecated_model_names (line 142) | def get_deprecated_model_names():
  function _process_mmcls_checkpoint (line 151) | def _process_mmcls_checkpoint(checkpoint):
  class CheckpointLoader (line 162) | class CheckpointLoader:
    method _register_scheme (line 168) | def _register_scheme(cls, prefixes, loader, force=False):
    method register_scheme (line 185) | def register_scheme(cls, prefixes, loader=None, force=False):
    method _get_checkpoint_loader (line 211) | def _get_checkpoint_loader(cls, path):
    method load_checkpoint (line 227) | def load_checkpoint(cls, filename, map_location=None, logger=None):
  function load_from_local (line 249) | def load_from_local(filename, map_location):
  function load_from_http (line 267) | def load_from_http(filename, map_location=None, model_dir=None):
  function load_from_pavi (line 295) | def load_from_pavi(filename, map_location=None):
  function load_from_ceph (line 327) | def load_from_ceph(filename, map_location=None, backend='petrel'):
  function load_from_torchvision (line 368) | def load_from_torchvision(filename, map_location=None):
  function load_from_openmmlab (line 391) | def load_from_openmmlab(filename, map_location=None):
  function load_from_mmcls (line 431) | def load_from_mmcls(filename, map_location=None):
  function _load_checkpoint (line 450) | def _load_checkpoint(filename, map_location=None, logger=None):
  function _load_checkpoint_with_prefix (line 470) | def _load_checkpoint_with_prefix(prefix, filename, map_location=None):
  function load_checkpoint (line 503) | def load_checkpoint(model,
  function weights_to_cpu (line 553) | def weights_to_cpu(state_dict):
  function _save_to_state_dict (line 570) | def _save_to_state_dict(module, destination, prefix, keep_vars):
  function get_state_dict (line 590) | def get_state_dict(module, destination=None, prefix='', keep_vars=False):
  function save_checkpoint (line 634) | def save_checkpoint(model,

FILE: lavis/common/annotator/uniformer/mmcv/runner/default_constructor.py
  class DefaultRunnerConstructor (line 5) | class DefaultRunnerConstructor:
    method __init__ (line 36) | def __init__(self, runner_cfg, default_args=None):
    method __call__ (line 43) | def __call__(self):

FILE: lavis/common/annotator/uniformer/mmcv/runner/dist_utils.py
  function init_dist (line 14) | def init_dist(launcher, backend='nccl', **kwargs):
  function _init_dist_pytorch (line 27) | def _init_dist_pytorch(backend, **kwargs):
  function _init_dist_mpi (line 35) | def _init_dist_mpi(backend, **kwargs):
  function _init_dist_slurm (line 43) | def _init_dist_slurm(backend, port=None):
  function get_dist_info (line 78) | def get_dist_info():
  function master_only (line 88) | def master_only(func):
  function allreduce_params (line 99) | def allreduce_params(params, coalesce=True, bucket_size_mb=-1):
  function allreduce_grads (line 121) | def allreduce_grads(params, coalesce=True, bucket_size_mb=-1):
  function _allreduce_coalesced (line 145) | def _allreduce_coalesced(tensors, world_size, bucket_size_mb=-1):

FILE: lavis/common/annotator/uniformer/mmcv/runner/epoch_based_runner.py
  class EpochBasedRunner (line 18) | class EpochBasedRunner(BaseRunner):
    method run_iter (line 24) | def run_iter(self, data_batch, train_mode, **kwargs):
    method train (line 40) | def train(self, data_loader, **kwargs):
    method val (line 58) | def val(self, data_loader, **kwargs):
    method run (line 72) | def run(self, data_loaders, workflow, max_epochs=None, **kwargs):
    method save_checkpoint (line 132) | def save_checkpoint(self,
  class Runner (line 181) | class Runner(EpochBasedRunner):
    method __init__ (line 184) | def __init__(self, *args, **kwargs):

FILE: lavis/common/annotator/uniformer/mmcv/runner/fp16_utils.py
  function cast_tensor_type (line 24) | def cast_tensor_type(inputs, src_type, dst_type):
  function auto_fp16 (line 55) | def auto_fp16(apply_to=None, out_fp32=False):
  function force_fp32 (line 141) | def force_fp32(apply_to=None, out_fp16=False):
  function allreduce_grads (line 227) | def allreduce_grads(params, coalesce=True, bucket_size_mb=-1):
  function wrap_fp16_model (line 234) | def wrap_fp16_model(model):
  function patch_norm_fp32 (line 263) | def patch_norm_fp32(module):
  function patch_forward_method (line 283) | def patch_forward_method(func, src_type, dst_type, convert_output=True):
  class LossScaler (line 306) | class LossScaler:
    method __init__ (line 335) | def __init__(self,
    method has_overflow (line 349) | def has_overflow(self, params):
    method _has_inf_or_nan (line 358) | def _has_inf_or_nan(x):
    method update_scale (line 372) | def update_scale(self, overflow):
    method state_dict (line 385) | def state_dict(self):
    method load_state_dict (line 395) | def load_state_dict(self, state_dict):
    method loss_scale (line 409) | def loss_scale(self):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/checkpoint.py
  class CheckpointHook (line 11) | class CheckpointHook(Hook):
    method __init__ (line 51) | def __init__(self,
    method before_run (line 71) | def before_run(self, runner):
    method after_train_epoch (line 102) | def after_train_epoch(self, runner):
    method _save_checkpoint (line 119) | def _save_checkpoint(self, runner):
    method after_train_iter (line 153) | def after_train_iter(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/closure.py
  class ClosureHook (line 6) | class ClosureHook(Hook):
    method __init__ (line 8) | def __init__(self, fn_name, fn):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/ema.py
  class EMAHook (line 7) | class EMAHook(Hook):
    method __init__ (line 29) | def __init__(self,
    method before_run (line 41) | def before_run(self, runner):
    method after_train_iter (line 60) | def after_train_iter(self, runner):
    method after_train_epoch (line 73) | def after_train_epoch(self, runner):
    method before_train_epoch (line 78) | def before_train_epoch(self, runner):
    method _swap_ema_parameters (line 83) | def _swap_ema_parameters(self):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/evaluation.py
  class EvalHook (line 16) | class EvalHook(Hook):
    method __init__ (line 85) | def __init__(self,
    method _init_rule (line 153) | def _init_rule(self, rule, key_indicator):
    method before_run (line 202) | def before_run(self, runner):
    method before_train_iter (line 228) | def before_train_iter(self, runner):
    method before_train_epoch (line 236) | def before_train_epoch(self, runner):
    method after_train_iter (line 244) | def after_train_iter(self, runner):
    method after_train_epoch (line 264) | def after_train_epoch(self, runner):
    method _do_evaluate (line 269) | def _do_evaluate(self, runner):
    method _should_evaluate (line 279) | def _should_evaluate(self, runner):
    method _save_ckpt (line 314) | def _save_ckpt(self, runner, key_score):
    method evaluate (line 354) | def evaluate(self, runner, results):
  class DistEvalHook (line 387) | class DistEvalHook(EvalHook):
    method __init__ (line 439) | def __init__(self,
    method _do_evaluate (line 478) | def _do_evaluate(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/hook.py
  class Hook (line 7) | class Hook:
    method before_run (line 13) | def before_run(self, runner):
    method after_run (line 16) | def after_run(self, runner):
    method before_epoch (line 19) | def before_epoch(self, runner):
    method after_epoch (line 22) | def after_epoch(self, runner):
    method before_iter (line 25) | def before_iter(self, runner):
    method after_iter (line 28) | def after_iter(self, runner):
    method before_train_epoch (line 31) | def before_train_epoch(self, runner):
    method before_val_epoch (line 34) | def before_val_epoch(self, runner):
    method after_train_epoch (line 37) | def after_train_epoch(self, runner):
    method after_val_epoch (line 40) | def after_val_epoch(self, runner):
    method before_train_iter (line 43) | def before_train_iter(self, runner):
    method before_val_iter (line 46) | def before_val_iter(self, runner):
    method after_train_iter (line 49) | def after_train_iter(self, runner):
    method after_val_iter (line 52) | def after_val_iter(self, runner):
    method every_n_epochs (line 55) | def every_n_epochs(self, runner, n):
    method every_n_inner_iters (line 58) | def every_n_inner_iters(self, runner, n):
    method every_n_iters (line 61) | def every_n_iters(self, runner, n):
    method end_of_epoch (line 64) | def end_of_epoch(self, runner):
    method is_last_epoch (line 67) | def is_last_epoch(self, runner):
    method is_last_iter (line 70) | def is_last_iter(self, runner):
    method get_triggered_stages (line 73) | def get_triggered_stages(self):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/iter_timer.py
  class IterTimerHook (line 8) | class IterTimerHook(Hook):
    method before_epoch (line 10) | def before_epoch(self, runner):
    method before_iter (line 13) | def before_iter(self, runner):
    method after_iter (line 16) | def after_iter(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/logger/base.py
  class LoggerHook (line 11) | class LoggerHook(Hook):
    method __init__ (line 24) | def __init__(self,
    method log (line 35) | def log(self, runner):
    method is_scalar (line 39) | def is_scalar(val, include_np=True, include_torch=True):
    method get_mode (line 59) | def get_mode(self, runner):
    method get_epoch (line 72) | def get_epoch(self, runner):
    method get_iter (line 84) | def get_iter(self, runner, inner_iter=False):
    method get_lr_tags (line 92) | def get_lr_tags(self, runner):
    method get_momentum_tags (line 102) | def get_momentum_tags(self, runner):
    method get_loggable_tags (line 112) | def get_loggable_tags(self,
    method before_run (line 133) | def before_run(self, runner):
    method before_epoch (line 139) | def before_epoch(self, runner):
    method after_train_iter (line 142) | def after_train_iter(self, runner):
    method after_train_epoch (line 156) | def after_train_epoch(self, runner):
    method after_val_epoch (line 162) | def after_val_epoch(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/logger/dvclive.py
  class DvcliveLoggerHook (line 8) | class DvcliveLoggerHook(LoggerHook):
    method __init__ (line 29) | def __init__(self,
    method import_dvclive (line 41) | def import_dvclive(self):
    method before_run (line 50) | def before_run(self, runner):
    method log (line 54) | def log(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/logger/mlflow.py
  class MlflowLoggerHook (line 8) | class MlflowLoggerHook(LoggerHook):
    method __init__ (line 10) | def __init__(self,
    method import_mlflow (line 51) | def import_mlflow(self):
    method before_run (line 62) | def before_run(self, runner):
    method log (line 70) | def log(self, runner):
    method after_run (line 76) | def after_run(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/logger/neptune.py
  class NeptuneLoggerHook (line 8) | class NeptuneLoggerHook(LoggerHook):
    method __init__ (line 38) | def __init__(self,
    method import_neptune (line 52) | def import_neptune(self):
    method before_run (line 62) | def before_run(self, runner):
    method log (line 69) | def log(self, runner):
    method after_run (line 81) | def after_run(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/logger/pavi.py
  class PaviLoggerHook (line 17) | class PaviLoggerHook(LoggerHook):
    method __init__ (line 19) | def __init__(self,
    method before_run (line 36) | def before_run(self, runner):
    method get_step (line 74) | def get_step(self, runner):
    method log (line 82) | def log(self, runner):
    method after_run (line 89) | def after_run(self, runner):
    method before_epoch (line 107) | def before_epoch(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/logger/tensorboard.py
  class TensorboardLoggerHook (line 11) | class TensorboardLoggerHook(LoggerHook):
    method __init__ (line 13) | def __init__(self,
    method before_run (line 24) | def before_run(self, runner):
    method log (line 47) | def log(self, runner):
    method after_run (line 56) | def after_run(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/logger/text.py
  class TextLoggerHook (line 18) | class TextLoggerHook(LoggerHook):
    method __init__ (line 55) | def __init__(self,
    method before_run (line 89) | def before_run(self, runner):
    method _get_max_memory (line 109) | def _get_max_memory(self, runner):
    method _log_info (line 119) | def _log_info(self, log_dict, runner):
    method _dump_log (line 185) | def _dump_log(self, log_dict, runner):
    method _round_float (line 196) | def _round_float(self, items):
    method log (line 204) | def log(self, runner):
    method after_run (line 238) | def after_run(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/logger/wandb.py
  class WandbLoggerHook (line 8) | class WandbLoggerHook(LoggerHook):
    method __init__ (line 10) | def __init__(self,
    method import_wandb (line 25) | def import_wandb(self):
    method before_run (line 34) | def before_run(self, runner):
    method log (line 44) | def log(self, runner):
    method after_run (line 55) | def after_run(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/lr_updater.py
  class LrUpdaterHook (line 9) | class LrUpdaterHook(Hook):
    method __init__ (line 25) | def __init__(self,
    method _set_lr (line 58) | def _set_lr(self, runner, lr_groups):
    method get_lr (line 68) | def get_lr(self, runner, base_lr):
    method get_regular_lr (line 71) | def get_regular_lr(self, runner):
    method get_warmup_lr (line 85) | def get_warmup_lr(self, cur_iters):
    method before_run (line 107) | def before_run(self, runner):
    method before_train_epoch (line 126) | def before_train_epoch(self, runner):
    method before_train_iter (line 137) | def before_train_iter(self, runner):
  class FixedLrUpdaterHook (line 157) | class FixedLrUpdaterHook(LrUpdaterHook):
    method __init__ (line 159) | def __init__(self, **kwargs):
    method get_lr (line 162) | def get_lr(self, runner, base_lr):
  class StepLrUpdaterHook (line 167) | class StepLrUpdaterHook(LrUpdaterHook):
    method __init__ (line 180) | def __init__(self, step, gamma=0.1, min_lr=None, **kwargs):
    method get_lr (line 193) | def get_lr(self, runner, base_lr):
  class ExpLrUpdaterHook (line 214) | class ExpLrUpdaterHook(LrUpdaterHook):
    method __init__ (line 216) | def __init__(self, gamma, **kwargs):
    method get_lr (line 220) | def get_lr(self, runner, base_lr):
  class PolyLrUpdaterHook (line 226) | class PolyLrUpdaterHook(LrUpdaterHook):
    method __init__ (line 228) | def __init__(self, power=1., min_lr=0., **kwargs):
    method get_lr (line 233) | def get_lr(self, runner, base_lr):
  class InvLrUpdaterHook (line 245) | class InvLrUpdaterHook(LrUpdaterHook):
    method __init__ (line 247) | def __init__(self, gamma, power=1., **kwargs):
    method get_lr (line 252) | def get_lr(self, runner, base_lr):
  class CosineAnnealingLrUpdaterHook (line 258) | class CosineAnnealingLrUpdaterHook(LrUpdaterHook):
    method __init__ (line 260) | def __init__(self, min_lr=None, min_lr_ratio=None, **kwargs):
    method get_lr (line 266) | def get_lr(self, runner, base_lr):
  class FlatCosineAnnealingLrUpdaterHook (line 282) | class FlatCosineAnnealingLrUpdaterHook(LrUpdaterHook):
    method __init__ (line 298) | def __init__(self,
    method get_lr (line 314) | def get_lr(self, runner, base_lr):
  class CosineRestartLrUpdaterHook (line 336) | class CosineRestartLrUpdaterHook(LrUpdaterHook):
    method __init__ (line 349) | def __init__(self,
    method get_lr (line 368) | def get_lr(self, runner, base_lr):
  function get_position_from_periods (line 388) | def get_position_from_periods(iteration, cumulative_periods):
  class CyclicLrUpdaterHook (line 412) | class CyclicLrUpdaterHook(LrUpdaterHook):
    method __init__ (line 434) | def __init__(self,
    method before_run (line 472) | def before_run(self, runner):
    method get_lr (line 485) | def get_lr(self, runner, base_lr):
  class OneCycleLrUpdaterHook (line 498) | class OneCycleLrUpdaterHook(LrUpdaterHook):
    method __init__ (line 532) | def __init__(self,
    method before_run (line 575) | def before_run(self, runner):
    method get_lr (line 614) | def get_lr(self, runner, base_lr):
  function annealing_cos (line 627) | def annealing_cos(start, end, factor, weight=1):
  function annealing_linear (line 645) | def annealing_linear(start, end, factor):
  function format_param (line 659) | def format_param(name, optim, param):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/memory.py
  class EmptyCacheHook (line 8) | class EmptyCacheHook(Hook):
    method __init__ (line 10) | def __init__(self, before_epoch=False, after_epoch=True, after_iter=Fa...
    method after_iter (line 15) | def after_iter(self, runner):
    method before_epoch (line 19) | def before_epoch(self, runner):
    method after_epoch (line 23) | def after_epoch(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/momentum_updater.py
  class MomentumUpdaterHook (line 7) | class MomentumUpdaterHook(Hook):
    method __init__ (line 9) | def __init__(self,
    method _set_momentum (line 35) | def _set_momentum(self, runner, momentum_groups):
    method get_momentum (line 52) | def get_momentum(self, runner, base_momentum):
    method get_regular_momentum (line 55) | def get_regular_momentum(self, runner):
    method get_warmup_momentum (line 71) | def get_warmup_momentum(self, cur_iters):
    method before_run (line 101) | def before_run(self, runner):
    method before_train_epoch (line 128) | def before_train_epoch(self, runner):
    method before_train_iter (line 134) | def before_train_iter(self, runner):
  class StepMomentumUpdaterHook (line 154) | class StepMomentumUpdaterHook(MomentumUpdaterHook):
    method __init__ (line 168) | def __init__(self, step, gamma=0.5, min_momentum=None, **kwargs):
    method get_momentum (line 181) | def get_momentum(self, runner, base_momentum):
  class CosineAnnealingMomentumUpdaterHook (line 202) | class CosineAnnealingMomentumUpdaterHook(MomentumUpdaterHook):
    method __init__ (line 204) | def __init__(self, min_momentum=None, min_momentum_ratio=None, **kwargs):
    method get_momentum (line 210) | def get_momentum(self, runner, base_momentum):
  class CyclicMomentumUpdaterHook (line 226) | class CyclicMomentumUpdaterHook(MomentumUpdaterHook):
    method __init__ (line 244) | def __init__(self,
    method before_run (line 273) | def before_run(self, runner):
    method get_momentum (line 286) | def get_momentum(self, runner, base_momentum):
  class OneCycleMomentumUpdaterHook (line 299) | class OneCycleMomentumUpdaterHook(MomentumUpdaterHook):
    method __init__ (line 333) | def __init__(self,
    method before_run (line 371) | def before_run(self, runner):
    method _set_momentum (line 448) | def _set_momentum(self, runner, momentum_groups):
    method get_momentum (line 465) | def get_momentum(self, runner, param_group):
    method get_regular_momentum (line 479) | def get_regular_momentum(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/optimizer.py
  class OptimizerHook (line 22) | class OptimizerHook(Hook):
    method __init__ (line 24) | def __init__(self, grad_clip=None):
    method clip_grads (line 27) | def clip_grads(self, params):
    method after_train_iter (line 33) | def after_train_iter(self, runner):
  class GradientCumulativeOptimizerHook (line 46) | class GradientCumulativeOptimizerHook(OptimizerHook):
    method __init__ (line 64) | def __init__(self, cumulative_iters=1, **kwargs):
    method has_batch_norm (line 76) | def has_batch_norm(self, module):
    method _init (line 84) | def _init(self, runner):
    method after_train_iter (line 105) | def after_train_iter(self, runner):
  class Fp16OptimizerHook (line 134) | class Fp16OptimizerHook(OptimizerHook):
    method __init__ (line 162) | def __init__(self,
    method before_run (line 184) | def before_run(self, runner):
    method copy_grads_to_fp32 (line 193) | def copy_grads_to_fp32(self, fp16_net, fp32_weights):
    method copy_params_to_fp16 (line 203) | def copy_params_to_fp16(self, fp16_net, fp32_weights):
    method after_train_iter (line 209) | def after_train_iter(self, runner):
    method __init__ (line 318) | def __init__(self,
    method before_run (line 339) | def before_run(self, runner):
    method copy_grads_to_fp32 (line 367) | def copy_grads_to_fp32(self, fp16_net, fp32_weights):
    method copy_params_to_fp16 (line 377) | def copy_params_to_fp16(self, fp16_net, fp32_weights):
    method after_train_iter (line 383) | def after_train_iter(self, runner):
  class GradientCumulativeFp16OptimizerHook (line 242) | class GradientCumulativeFp16OptimizerHook(GradientCumulativeOptimizerHook,
    method __init__ (line 251) | def __init__(self, *args, **kwargs):
    method after_train_iter (line 255) | def after_train_iter(self, runner):
    method __init__ (line 444) | def __init__(self, *args, **kwargs):
    method after_train_iter (line 448) | def after_train_iter(self, runner):
  class Fp16OptimizerHook (line 297) | class Fp16OptimizerHook(OptimizerHook):
    method __init__ (line 162) | def __init__(self,
    method before_run (line 184) | def before_run(self, runner):
    method copy_grads_to_fp32 (line 193) | def copy_grads_to_fp32(self, fp16_net, fp32_weights):
    method copy_params_to_fp16 (line 203) | def copy_params_to_fp16(self, fp16_net, fp32_weights):
    method after_train_iter (line 209) | def after_train_iter(self, runner):
    method __init__ (line 318) | def __init__(self,
    method before_run (line 339) | def before_run(self, runner):
    method copy_grads_to_fp32 (line 367) | def copy_grads_to_fp32(self, fp16_net, fp32_weights):
    method copy_params_to_fp16 (line 377) | def copy_params_to_fp16(self, fp16_net, fp32_weights):
    method after_train_iter (line 383) | def after_train_iter(self, runner):
  class GradientCumulativeFp16OptimizerHook (line 439) | class GradientCumulativeFp16OptimizerHook(GradientCumulativeOptimizerHook,
    method __init__ (line 251) | def __init__(self, *args, **kwargs):
    method after_train_iter (line 255) | def after_train_iter(self, runner):
    method __init__ (line 444) | def __init__(self, *args, **kwargs):
    method after_train_iter (line 448) | def after_train_iter(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/profiler.py
  class ProfilerHook (line 12) | class ProfilerHook(Hook):
    method __init__ (line 55) | def __init__(self,
    method before_run (line 107) | def before_run(self, runner):
    method after_train_epoch (line 166) | def after_train_epoch(self, runner):
    method after_train_iter (line 174) | def after_train_iter(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/sampler_seed.py
  class DistSamplerSeedHook (line 6) | class DistSamplerSeedHook(Hook):
    method before_epoch (line 14) | def before_epoch(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/hooks/sync_buffer.py
  class SyncBuffersHook (line 7) | class SyncBuffersHook(Hook):
    method __init__ (line 16) | def __init__(self, distributed=True):
    method after_epoch (line 19) | def after_epoch(self, runner):

FILE: lavis/common/annotator/uniformer/mmcv/runner/iter_based_runner.py
  class IterLoader (line 19) | class IterLoader:
    method __init__ (line 21) | def __init__(self, dataloader):
    method epoch (line 27) | def epoch(self):
    method __next__ (line 30) | def __next__(self):
    method __len__ (line 43) | def __len__(self):
  class IterBasedRunner (line 48) | class IterBasedRunner(BaseRunner):
    method train (line 54) | def train(self, data_loader, **kwargs):
    method val (line 72) | def val(self, data_loader, **kwargs):
    method run (line 87) | def run(self, data_loaders, workflow, max_iters=None, **kwargs):
    method resume (line 140) | def resume(self,
    method save_checkpoint (line 179) | def save_checkpoint(self,
    method register_training_hooks (line 224) | def register_training_hooks(self,

FILE: lavis/common/annotator/uniformer/mmcv/runner/log_buffer.py
  class LogBuffer (line 7) | class LogBuffer:
    method __init__ (line 9) | def __init__(self):
    method clear (line 15) | def clear(self):
    method clear_output (line 20) | def clear_output(self):
    method update (line 24) | def update(self, vars, count=1):
    method average (line 33) | def average(self, n=0):

FILE: lavis/common/annotator/uniformer/mmcv/runner/optimizer/builder.py
  function register_torch_optimizers (line 13) | def register_torch_optimizers():
  function build_optimizer_constructor (line 29) | def build_optimizer_constructor(cfg):
  function build_optimizer (line 33) | def build_optimizer(model, cfg):

FILE: lavis/common/annotator/uniformer/mmcv/runner/optimizer/default_constructor.py
  class DefaultOptimizerConstructor (line 13) | class DefaultOptimizerConstructor:
    method __init__ (line 95) | def __init__(self, optimizer_cfg, paramwise_cfg=None):
    method _validate_cfg (line 105) | def _validate_cfg(self):
    method _is_in (line 128) | def _is_in(self, param_group, param_group_list):
    method add_params (line 137) | def add_params(self, params, module, prefix='', is_dcn_module=None):
    method __call__ (line 234) | def __call__(self, model):

FILE: lavis/common/annotator/uniformer/mmcv/runner/priority.py
  class Priority (line 5) | class Priority(Enum):
  function get_priority (line 42) | def get_priority(priority):

FILE: lavis/common/annotator/uniformer/mmcv/runner/utils.py
  function get_host_info (line 16) | def get_host_info():
  function get_time_str (line 31) | def get_time_str():
  function obj_from_dict (line 35) | def obj_from_dict(info, parent=None, default_args=None):
  function set_random_seed (line 70) | def set_random_seed(seed, deterministic=False, use_rank_shift=False):

FILE: lavis/common/annotator/uniformer/mmcv/utils/config.py
  class ConfigDict (line 33) | class ConfigDict(Dict):
    method __missing__ (line 35) | def __missing__(self, name):
    method __getattr__ (line 38) | def __getattr__(self, name):
  function add_args (line 51) | def add_args(parser, cfg, prefix=''):
  class Config (line 70) | class Config:
    method _validate_py_syntax (line 96) | def _validate_py_syntax(filename):
    method _substitute_predefined_vars (line 107) | def _substitute_predefined_vars(filename, temp_config_name):
    method _pre_substitute_base_vars (line 128) | def _pre_substitute_base_vars(filename, temp_config_name):
    method _substitute_base_vars (line 147) | def _substitute_base_vars(cfg, base_var_dict, base_cfg):
    method _file2dict (line 179) | def _file2dict(filename, use_predefined_variables=True):
    method _merge_a_into_b (line 274) | def _merge_a_into_b(a, b, allow_list_keys=False):
    method fromfile (line 328) | def fromfile(filename,
    method fromstring (line 338) | def fromstring(cfg_str, file_format):
    method auto_argparser (line 366) | def auto_argparser(description=None):
    method __init__ (line 377) | def __init__(self, cfg_dict=None, cfg_text=None, filename=None):
    method filename (line 399) | def filename(self):
    method text (line 403) | def text(self):
    method pretty_text (line 407) | def pretty_text(self):
    method __repr__ (line 500) | def __repr__(self):
    method __len__ (line 503) | def __len__(self):
    method __getattr__ (line 506) | def __getattr__(self, name):
    method __getitem__ (line 509) | def __getitem__(self, name):
    method __setattr__ (line 512) | def __setattr__(self, name, value):
    method __setitem__ (line 517) | def __setitem__(self, name, value):
    method __iter__ (line 522) | def __iter__(self):
    method __getstate__ (line 525) | def __getstate__(self):
    method __setstate__ (line 528) | def __setstate__(self, state):
    method dump (line 534) | def dump(self, file=None):
    method merge_from_dict (line 550) | def merge_from_dict(self, options, allow_list_keys=True):
  class DictAction (line 597) | class DictAction(Action):
    method _parse_int_float_bool (line 607) | def _parse_int_float_bool(val):
    method _parse_iterable (line 621) | def _parse_iterable(val):
    method __call__ (line 683) | def __call__(self, parser, namespace, values, option_string=None):

FILE: lavis/common/annotator/uniformer/mmcv/utils/env.py
  function collect_env (line 16) | def collect_env():

FILE: lavis/common/annotator/uniformer/mmcv/utils/ext_loader.py
  function load_ext (line 12) | def load_ext(name, funcs):
  function get_fake_func (line 41) | def get_fake_func(name, e):
  function load_ext (line 49) | def load_ext(name, funcs):
  function check_ops_exist (line 69) | def check_ops_exist():

FILE: lavis/common/annotator/uniformer/mmcv/utils/logging.py
  function get_logger (line 9) | def get_logger(name, log_file=None, log_level=logging.INFO, file_mode='w'):
  function print_log (line 85) | def print_log(msg, logger=None, level=logging.INFO):

FILE: lavis/common/annotator/uniformer/mmcv/utils/misc.py
  function _ntuple (line 14) | def _ntuple(n):
  function is_str (line 31) | def is_str(x):
  function import_modules_from_strings (line 39) | def import_modules_from_strings(imports, allow_failed_imports=False):
  function iter_cast (line 87) | def iter_cast(inputs, dst_type, return_type=None):
  function list_cast (line 112) | def list_cast(inputs, dst_type):
  function tuple_cast (line 120) | def tuple_cast(inputs, dst_type):
  function is_seq_of (line 128) | def is_seq_of(seq, expected_type, seq_type=None):
  function is_list_of (line 152) | def is_list_of(seq, expected_type):
  function is_tuple_of (line 160) | def is_tuple_of(seq, expected_type):
  function slice_list (line 168) | def slice_list(in_list, lens):
  function concat_list (line 194) | def concat_list(in_list):
  function check_prerequisites (line 206) | def check_prerequisites(
  function _check_py_package (line 244) | def _check_py_package(package):
  function _check_executable (line 253) | def _check_executable(cmd):
  function requires_package (line 260) | def requires_package(prerequisites):
  function requires_executable (line 276) | def requires_executable(prerequisites):
  function deprecated_api_warning (line 288) | def deprecated_api_warning(name_dict, cls_name=None):
  function is_method_overridden (line 348) | def is_method_overridden(method, base_class, derived_class):
  function has_method (line 367) | def has_method(obj: object, method: str) -> bool:

FILE: lavis/common/annotator/uniformer/mmcv/utils/parrots_jit.py
  function jit (line 12) | def jit(func=None,
  function skip_no_elena (line 36) | def skip_no_elena(func):

FILE: lavis/common/annotator/uniformer/mmcv/utils/parrots_wrapper.py
  function is_rocm_pytorch (line 9) | def is_rocm_pytorch() -> bool:
  function _get_cuda_home (line 21) | def _get_cuda_home():
  function get_build_config (line 33) | def get_build_config():
  function _get_conv (line 41) | def _get_conv():
  function _get_dataloader (line 49) | def _get_dataloader():
  function _get_extension (line 58) | def _get_extension():
  function _get_pool (line 69) | def _get_pool():
  function _get_norm (line 81) | def _get_norm():
  class SyncBatchNorm (line 99) | class SyncBatchNorm(SyncBatchNorm_):
    method _check_input_dim (line 101) | def _check_input_dim(self, input):

FILE: lavis/common/annotator/uniformer/mmcv/utils/path.py
  function is_filepath (line 9) | def is_filepath(x):
  function fopen (line 13) | def fopen(filepath, *args, **kwargs):
  function check_file_exist (line 21) | def check_file_exist(filename, msg_tmpl='file "{}" does not exist'):
  function mkdir_or_exist (line 26) | def mkdir_or_exist(dir_name, mode=0o777):
  function symlink (line 33) | def symlink(src, dst, overwrite=True, **kwargs):
  function scandir (line 39) | def scandir(dir_path, suffix=None, recursive=False, case_sensitive=True):
  function find_vcs_root (line 83) | def find_vcs_root(path, markers=('.git', )):

FILE: lavis/common/annotator/uniformer/mmcv/utils/progressbar.py
  class ProgressBar (line 10) | class ProgressBar:
    method __init__ (line 13) | def __init__(self, task_num=0, bar_width=50, start=True, file=sys.stdo...
    method terminal_width (line 22) | def terminal_width(self):
    method start (line 26) | def start(self):
    method update (line 35) | def update(self, num_tasks=1):
  function track_progress (line 64) | def track_progress(func, tasks, bar_width=50, file=sys.stdout, **kwargs):
  function init_pool (line 98) | def init_pool(process_num, initializer=None, initargs=None):
  function track_parallel_progress (line 109) | def track_parallel_progress(func,
  function track_iter_progress (line 179) | def track_iter_progress(tasks, bar_width=50, file=sys.stdout):

FILE: lavis/common/annotator/uniformer/mmcv/utils/registry.py
  function build_from_cfg (line 9) | def build_from_cfg(cfg, registry, default_args=None):
  class Registry (line 58) | class Registry:
    method __init__ (line 88) | def __init__(self, name, build_func=None, parent=None, scope=None):
    method __len__ (line 112) | def __len__(self):
    method __contains__ (line 115) | def __contains__(self, key):
    method __repr__ (line 118) | def __repr__(self):
    method infer_scope (line 125) | def infer_scope():
    method split_scope_key (line 149) | def split_scope_key(key):
    method name (line 171) | def name(self):
    method scope (line 175) | def scope(self):
    method module_dict (line 179) | def module_dict(self):
    method children (line 183) | def children(self):
    method get (line 186) | def get(self, key):
    method build (line 211) | def build(self, *args, **kwargs):
    method _add_children (line 214) | def _add_children(self, registry):
    method _register_module (line 235) | def _register_module(self, module_class, module_name=None, force=False):
    method deprecated_register_module (line 250) | def deprecated_register_module(self, cls=None, force=False):
    method register_module (line 260) | def register_module(self, name=None, force=False, module=None):

FILE: lavis/common/annotator/uniformer/mmcv/utils/testing.py
  function check_python_script (line 10) | def check_python_script(cmd):
  function _any (line 25) | def _any(judge_result):
  function assert_dict_contains_subset (line 42) | def assert_dict_contains_subset(dict_obj: Dict[Any, Any],
  function assert_attrs_equal (line 61) | def assert_attrs_equal(obj: Any, expected_attrs: Dict[str, Any]) -> bool:
  function assert_dict_has_keys (line 77) | def assert_dict_has_keys(obj: Dict[str, Any],
  function assert_keys_equal (line 92) | def assert_keys_equal(result_keys: List[str], target_keys: List[str]) ->...
  function assert_is_norm_layer (line 105) | def assert_is_norm_layer(module) -> bool:
  function assert_params_all_zeros (line 120) | def assert_params_all_zeros(module) -> bool:

FILE: lavis/common/annotator/uniformer/mmcv/utils/timer.py
  class TimerError (line 5) | class TimerError(Exception):
    method __init__ (line 7) | def __init__(self, message):
  class Timer (line 12) | class Timer:
    method __init__ (line 38) | def __init__(self, start=True, print_tmpl=None):
    method is_running (line 45) | def is_running(self):
    method __enter__ (line 49) | def __enter__(self):
    method __exit__ (line 53) | def __exit__(self, type, value, traceback):
    method start (line 57) | def start(self):
    method since_start (line 64) | def since_start(self):
    method since_last_check (line 74) | def since_last_check(self):
  function check_time (line 92) | def check_time(timer_id):

FILE: lavis/common/annotator/uniformer/mmcv/utils/trace.py
  function is_jit_tracing (line 8) | def is_jit_tracing() -> bool:

FILE: lavis/common/annotator/uniformer/mmcv/utils/version_utils.py
  function digit_version (line 9) | def digit_version(version_str: str, length: int = 4):
  function _minimal_ext_cmd (line 50) | def _minimal_ext_cmd(cmd):
  function get_git_hash (line 66) | def get_git_hash(fallback='unknown', digits=None):

FILE: lavis/common/annotator/uniformer/mmcv/version.py
  function parse_version_info (line 5) | def parse_version_info(version_str: str, length: int = 4) -> tuple:

FILE: lavis/common/annotator/uniformer/mmcv/video/io.py
  class Cache (line 14) | class Cache:
    method __init__ (line 16) | def __init__(self, capacity):
    method capacity (line 23) | def capacity(self):
    method size (line 27) | def size(self):
    method put (line 30) | def put(self, key, val):
    method get (line 37) | def get(self, key, default=None):
  class VideoReader (line 42) | class VideoReader:
    method __init__ (line 64) | def __init__(self, filename, cache_capacity=10):
    method vcap (line 80) | def vcap(self):
    method opened (line 85) | def opened(self):
    method width (line 90) | def width(self):
    method height (line 95) | def height(self):
    method resolution (line 100) | def resolution(self):
    method fps (line 105) | def fps(self):
    method frame_cnt (line 110) | def frame_cnt(self):
    method fourcc (line 115) | def fourcc(self):
    method position (line 120) | def position(self):
    method _get_real_position (line 124) | def _get_real_position(self):
    method _set_real_position (line 127) | def _set_real_position(self, frame_id):
    method read (line 134) | def read(self):
    method get_frame (line 160) | def get_frame(self, frame_id):
    method current_frame (line 187) | def current_frame(self):
    method cvt2frames (line 198) | def cvt2frames(self,
    method __len__ (line 240) | def __len__(self):
    method __getitem__ (line 243) | def __getitem__(self, index):
    method __iter__ (line 256) | def __iter__(self):
    method __next__ (line 260) | def __next__(self):
    method __enter__ (line 269) | def __enter__(self):
    method __exit__ (line 272) | def __exit__(self, exc_type, exc_value, traceback):
  function frames2video (line 276) | def frames2video(frame_dir,

FILE: lavis/common/annotator/uniformer/mmcv/video/optflow.py
  function flowread (line 12) | def flowread(flow_or_path, quantize=False, concat_axis=0, *args, **kwargs):
  function flowwrite (line 61) | def flowwrite(flow, filename, quantize=False, concat_axis=0, *args, **kw...
  function quantize_flow (line 91) | def quantize_flow(flow, max_val=0.02, norm=True):
  function dequantize_flow (line 119) | def dequantize_flow(dx, dy, max_val=0.02, denorm=True):
  function flow_warp (line 143) | def flow_warp(img, flow, filling_value=0, interpolate_mode='nearest'):
  function flow_from_bytes (line 204) | def flow_from_bytes(content):
  function sparse_flow_from_bytes (line 234) | def sparse_flow_from_bytes(content):

FILE: lavis/common/annotator/uniformer/mmcv/video/processing.py
  function convert_video (line 11) | def convert_video(in_file,
  function resize_video (line 55) | def resize_video(in_file,
  function cut_video (line 93) | def cut_video(in_file,
  function concat_video (line 128) | def concat_video(video_list,

FILE: lavis/common/annotator/uniformer/mmcv/visualization/color.py
  class Color (line 9) | class Color(Enum):
  function color_val (line 24) | def color_val(color):

FILE: lavis/common/annotator/uniformer/mmcv/visualization/image.py
  function imshow (line 9) | def imshow(img, win_name='', wait_time=0):
  function imshow_bboxes (line 30) | def imshow_bboxes(img,
  function imshow_det_bboxes (line 84) | def imshow_det_bboxes(img,

FILE: lavis/common/annotator/uniformer/mmcv/visualization/optflow.py
  function flowshow (line 11) | def flowshow(flow, win_name='', wait_time=0):
  function flow2rgb (line 24) | def flow2rgb(flow, color_wheel=None, unknown_thr=1e6):
  function make_color_wheel (line 76) | def make_color_wheel(bins=None):

FILE: lavis/common/annotator/uniformer/mmcv_custom/checkpoint.py
  function _get_mmcv_home (line 30) | def _get_mmcv_home():
  function load_state_dict (line 41) | def load_state_dict(module, state_dict, strict=False, logger=None):
  function load_url_dist (line 109) | def load_url_dist(url, model_dir=None):
  function load_pavimodel_dist (line 123) | def load_pavimodel_dist(model_path, map_location=None):
  function load_fileclient_dist (line 151) | def load_fileclient_dist(filename, backend, map_location):
  function get_torchvision_models (line 172) | def get_torchvision_models():
  function get_external_models (line 184) | def get_external_models():
  function get_mmcls_models (line 198) | def get_mmcls_models():
  function get_deprecated_model_names (line 205) | def get_deprecated_model_names():
  function _process_mmcls_checkpoint (line 214) | def _process_mmcls_checkpoint(checkpoint):
  function _load_checkpoint (line 225) | def _load_checkpoint(filename, map_location=None):
  function load_checkpoint (line 286) | def load_checkpoint(model,
  function weights_to_cpu (line 359) | def weights_to_cpu(state_dict):
  function _save_to_state_dict (line 374) | def _save_to_state_dict(module, destination, prefix, keep_vars):
  function get_state_dict (line 394) | def get_state_dict(module, destination=None, prefix='', keep_vars=False):
  function save_checkpoint (line 438) | def save_checkpoint(model, filename, optimizer=None, meta=None):

FILE: lavis/common/annotator/uniformer/mmseg/apis/inference.py
  function init_segmentor (line 11) | def init_segmentor(config, checkpoint=None, device='cuda:0'):
  class LoadImage (line 42) | class LoadImage:
    method __call__ (line 45) | def __call__(self, results):
  function inference_segmentor (line 69) | def inference_segmentor(model, img):
  function show_result_pyplot (line 101) | def show_result_pyplot(model,

FILE: lavis/common/annotator/uniformer/mmseg/apis/test.py
  function np2tmp (line 14) | def np2tmp(array, temp_file_name=None):
  function single_gpu_test (line 34) | def single_gpu_test(model,
  function multi_gpu_test (line 106) | def multi_gpu_test(model,
  function collect_results_cpu (line 164) | def collect_results_cpu(result_part, size, tmpdir=None):
  function collect_results_gpu (line 207) | def collect_results_gpu(result_part, size):

FILE: lavis/common/annotator/uniformer/mmseg/apis/train.py
  function set_random_seed (line 14) | def set_random_seed(seed, deterministic=False):
  function train_segmentor (line 33) | def train_segmentor(model,

FILE: lavis/common/annotator/uniformer/mmseg/core/evaluation/class_names.py
  function cityscapes_classes (line 4) | def cityscapes_classes():
  function ade_classes (line 14) | def ade_classes():
  function voc_classes (line 44) | def voc_classes():
  function cityscapes_palette (line 54) | def cityscapes_palette():
  function ade_palette (line 63) | def ade_palette():
  function voc_palette (line 105) | def voc_palette():
  function get_classes (line 121) | def get_classes(dataset):
  function get_palette (line 138) | def get_palette(dataset):

FILE: lavis/common/annotator/uniformer/mmseg/core/evaluation/eval_hooks.py
  class EvalHook (line 7) | class EvalHook(_EvalHook):
    method __init__ (line 22) | def __init__(self, *args, by_epoch=False, efficient_test=False, **kwar...
    method after_train_iter (line 26) | def after_train_iter(self, runner):
    method after_train_epoch (line 42) | def after_train_epoch(self, runner):
  class DistEvalHook (line 55) | class DistEvalHook(_DistEvalHook):
    method __init__ (line 70) | def __init__(self, *args, by_epoch=False, efficient_test=False, **kwar...
    method after_train_iter (line 74) | def after_train_iter(self, runner):
    method after_train_epoch (line 93) | def after_train_epoch(self, runner):

FILE: lavis/common/annotator/uniformer/mmseg/core/evaluation/metrics.py
  function f_score (line 8) | def f_score(precision, recall, beta=1):
  function intersect_and_union (line 25) | def intersect_and_union(pred_label,
  function total_intersect_and_union (line 88) | def total_intersect_and_union(results,
  function mean_iou (line 133) | def mean_iou(results,
  function mean_dice (line 172) | def mean_dice(results,
  function mean_fscore (line 212) | def mean_fscore(results,
  function eval_metrics (line 257) | def eval_metrics(results,

FILE: lavis/common/annotator/uniformer/mmseg/core/seg/builder.py
  function build_pixel_sampler (line 6) | def build_pixel_sampler(cfg, **default_args):

FILE: lavis/common/annotator/uniformer/mmseg/core/seg/sampler/base_pixel_sampler.py
  class BasePixelSampler (line 4) | class BasePixelSampler(metaclass=ABCMeta):
    method __init__ (line 7) | def __init__(self, **kwargs):
    method sample (line 11) | def sample(self, seg_logit, seg_label):

FILE: lavis/common/annotator/uniformer/mmseg/core/seg/sampler/ohem_pixel_sampler.py
  class OHEMPixelSampler (line 9) | class OHEMPixelSampler(BasePixelSampler):
    method __init__ (line 23) | def __init__(self, context, thresh=None, min_kept=100000):
    method sample (line 30) | def sample(self, seg_logit, seg_label):

FILE: lavis/common/annotator/uniformer/mmseg/core/utils/misc.py
  function add_prefix (line 1) | def add_prefix(inputs, prefix):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/ade.py
  class ADE20KDataset (line 6) | class ADE20KDataset(CustomDataset):
    method __init__ (line 79) | def __init__(self, **kwargs):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/builder.py
  function _concat_dataset (line 25) | def _concat_dataset(cfg, default_args=None):
  function build_dataset (line 61) | def build_dataset(cfg, default_args=None):
  function build_dataloader (line 78) | def build_dataloader(dataset,
  function worker_init_fn (line 155) | def worker_init_fn(worker_id, num_workers, rank, seed):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/chase_db1.py
  class ChaseDB1Dataset (line 8) | class ChaseDB1Dataset(CustomDataset):
    method __init__ (line 21) | def __init__(self, **kwargs):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/cityscapes.py
  class CityscapesDataset (line 14) | class CityscapesDataset(CustomDataset):
    method __init__ (line 32) | def __init__(self, **kwargs):
    method _convert_to_label_id (line 39) | def _convert_to_label_id(result):
    method results2img (line 50) | def results2img(self, results, imgfile_prefix, to_label_id):
    method format_results (line 91) | def format_results(self, results, imgfile_prefix=None, to_label_id=True):
    method evaluate (line 124) | def evaluate(self,
    method _evaluate_cityscapes (line 164) | def _evaluate_cityscapes(self, results, logger, imgfile_prefix):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/custom.py
  class CustomDataset (line 19) | class CustomDataset(Dataset):
    method __init__ (line 75) | def __init__(self,
    method __len__ (line 116) | def __len__(self):
    method load_annotations (line 120) | def load_annotations(self, img_dir, img_suffix, ann_dir, seg_map_suffix,
    method get_ann_info (line 158) | def get_ann_info(self, idx):
    method pre_pipeline (line 170) | def pre_pipeline(self, results):
    method __getitem__ (line 178) | def __getitem__(self, idx):
    method prepare_train_img (line 194) | def prepare_train_img(self, idx):
    method prepare_test_img (line 211) | def prepare_test_img(self, idx):
    method format_results (line 227) | def format_results(self, results, **kwargs):
    method get_gt_seg_maps (line 230) | def get_gt_seg_maps(self, efficient_test=False):
    method get_classes_and_palette (line 243) | def get_classes_and_palette(self, classes=None, palette=None):
    method get_palette_for_custom_classes (line 287) | def get_palette_for_custom_classes(self, class_names, palette=None):
    method evaluate (line 306) | def evaluate(self,

FILE: lavis/common/annotator/uniformer/mmseg/datasets/dataset_wrappers.py
  class ConcatDataset (line 7) | class ConcatDataset(_ConcatDataset):
    method __init__ (line 17) | def __init__(self, datasets):
  class RepeatDataset (line 24) | class RepeatDataset(object):
    method __init__ (line 37) | def __init__(self, dataset, times):
    method __getitem__ (line 44) | def __getitem__(self, idx):
    method __len__ (line 48) | def __len__(self):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/drive.py
  class DRIVEDataset (line 8) | class DRIVEDataset(CustomDataset):
    method __init__ (line 21) | def __init__(self, **kwargs):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/hrf.py
  class HRFDataset (line 8) | class HRFDataset(CustomDataset):
    method __init__ (line 21) | def __init__(self, **kwargs):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/pascal_context.py
  class PascalContextDataset (line 8) | class PascalContextDataset(CustomDataset):
    method __init__ (line 47) | def __init__(self, split, **kwargs):
  class PascalContextDataset59 (line 58) | class PascalContextDataset59(CustomDataset):
    method __init__ (line 96) | def __init__(self, split, **kwargs):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/pipelines/compose.py
  class Compose (line 9) | class Compose(object):
    method __init__ (line 17) | def __init__(self, transforms):
    method __call__ (line 29) | def __call__(self, data):
    method __repr__ (line 45) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/pipelines/formating.py
  function to_tensor (line 11) | def to_tensor(data):
  class ToTensor (line 37) | class ToTensor(object):
    method __init__ (line 44) | def __init__(self, keys):
    method __call__ (line 47) | def __call__(self, results):
    method __repr__ (line 62) | def __repr__(self):
  class ImageToTensor (line 67) | class ImageToTensor(object):
    method __init__ (line 78) | def __init__(self, keys):
    method __call__ (line 81) | def __call__(self, results):
    method __repr__ (line 100) | def __repr__(self):
  class Transpose (line 105) | class Transpose(object):
    method __init__ (line 113) | def __init__(self, keys, order):
    method __call__ (line 117) | def __call__(self, results):
    method __repr__ (line 133) | def __repr__(self):
  class ToDataContainer (line 139) | class ToDataContainer(object):
    method __init__ (line 150) | def __init__(self,
    method __call__ (line 155) | def __call__(self, results):
    method __repr__ (line 173) | def __repr__(self):
  class DefaultFormatBundle (line 178) | class DefaultFormatBundle(object):
    method __call__ (line 189) | def __call__(self, results):
    method __repr__ (line 214) | def __repr__(self):
  class Collect (line 219) | class Collect(object):
    method __init__ (line 256) | def __init__(self,
    method __call__ (line 264) | def __call__(self, results):
    method __repr__ (line 286) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/pipelines/loading.py
  class LoadImageFromFile (line 10) | class LoadImageFromFile(object):
    method __init__ (line 31) | def __init__(self,
    method __call__ (line 42) | def __call__(self, results):
    method __repr__ (line 81) | def __repr__(self):
  class LoadAnnotations (line 90) | class LoadAnnotations(object):
    method __init__ (line 104) | def __init__(self,
    method __call__ (line 113) | def __call__(self, results):
    method __repr__ (line 149) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/pipelines/test_time_aug.py
  class MultiScaleFlipAug (line 10) | class MultiScaleFlipAug(object):
    method __init__ (line 53) | def __init__(self,
    method __call__ (line 93) | def __call__(self, results):
    method __repr__ (line 128) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/pipelines/transforms.py
  class Resize (line 10) | class Resize(object):
    method __init__ (line 41) | def __init__(self,
    method random_select (line 68) | def random_select(img_scales):
    method random_sample (line 86) | def random_sample(img_scales):
    method random_sample_ratio (line 113) | def random_sample_ratio(img_scale, ratio_range):
    method _random_scale (line 139) | def _random_scale(self, results):
    method _resize_img (line 177) | def _resize_img(self, results):
    method _resize_seg (line 199) | def _resize_seg(self, results):
    method __call__ (line 210) | def __call__(self, results):
    method __repr__ (line 228) | def __repr__(self):
  class RandomFlip (line 238) | class RandomFlip(object):
    method __init__ (line 252) | def __init__(self, prob=None, direction='horizontal'):
    method __call__ (line 259) | def __call__(self, results):
    method __repr__ (line 288) | def __repr__(self):
  class Pad (line 293) | class Pad(object):
    method __init__ (line 308) | def __init__(self,
    method _pad_img (line 321) | def _pad_img(self, results):
    method _pad_seg (line 334) | def _pad_seg(self, results):
    method __call__ (line 342) | def __call__(self, results):
    method __repr__ (line 356) | def __repr__(self):
  class Normalize (line 364) | class Normalize(object):
    method __init__ (line 376) | def __init__(self, mean, std, to_rgb=True):
    method __call__ (line 381) | def __call__(self, results):
    method __repr__ (line 398) | def __repr__(self):
  class Rerange (line 406) | class Rerange(object):
    method __init__ (line 416) | def __init__(self, min_value=0, max_value=255):
    method __call__ (line 423) | def __call__(self, results):
    method __repr__ (line 445) | def __repr__(self):
  class CLAHE (line 452) | class CLAHE(object):
    method __init__ (line 465) | def __init__(self, clip_limit=40.0, tile_grid_size=(8, 8)):
    method __call__ (line 472) | def __call__(self, results):
    method __repr__ (line 489) | def __repr__(self):
  class RandomCrop (line 497) | class RandomCrop(object):
    method __init__ (line 506) | def __init__(self, crop_size, cat_max_ratio=1., ignore_index=255):
    method get_crop_bbox (line 512) | def get_crop_bbox(self, img):
    method crop (line 523) | def crop(self, img, crop_bbox):
    method __call__ (line 529) | def __call__(self, results):
    method __repr__ (line 565) | def __repr__(self):
  class RandomRotate (line 570) | class RandomRotate(object):
    method __init__ (line 588) | def __init__(self,
    method __call__ (line 609) | def __call__(self, results):
    method __repr__ (line 641) | def __repr__(self):
  class RGB2Gray (line 653) | class RGB2Gray(object):
    method __init__ (line 668) | def __init__(self, out_channels=None, weights=(0.299, 0.587, 0.114)):
    method __call__ (line 676) | def __call__(self, results):
    method __repr__ (line 700) | def __repr__(self):
  class AdjustGamma (line 708) | class AdjustGamma(object):
    method __init__ (line 716) | def __init__(self, gamma=1.0):
    method __call__ (line 724) | def __call__(self, results):
    method __repr__ (line 739) | def __repr__(self):
  class SegRescale (line 744) | class SegRescale(object):
    method __init__ (line 751) | def __init__(self, scale_factor=1):
    method __call__ (line 754) | def __call__(self, results):
    method __repr__ (line 769) | def __repr__(self):
  class PhotoMetricDistortion (line 774) | class PhotoMetricDistortion(object):
    method __init__ (line 794) | def __init__(self,
    method convert (line 804) | def convert(self, img, alpha=1, beta=0):
    method brightness (line 810) | def brightness(self, img):
    method contrast (line 819) | def contrast(self, img):
    method saturation (line 827) | def saturation(self, img):
    method hue (line 838) | def hue(self, img):
    method __call__ (line 848) | def __call__(self, results):
    method __repr__ (line 881) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/stare.py
  class STAREDataset (line 8) | class STAREDataset(CustomDataset):
    method __init__ (line 21) | def __init__(self, **kwargs):

FILE: lavis/common/annotator/uniformer/mmseg/datasets/voc.py
  class PascalVOCDataset (line 8) | class PascalVOCDataset(CustomDataset):
    method __init__ (line 26) | def __init__(self, split, **kwargs):

FILE: lavis/common/annotator/uniformer/mmseg/models/backbones/cgnet.py
  class GlobalContextExtractor (line 13) | class GlobalContextExtractor(nn.Module):
    method __init__ (line 26) | def __init__(self, channel, reduction=16, with_cp=False):
    method forward (line 37) | def forward(self, x):
  class ContextGuidedBlock (line 53) | class ContextGuidedBlock(nn.Module):
    method __init__ (line 78) | def __init__(self,
    method forward (line 142) | def forward(self, x):
  class InputInjection (line 170) | class InputInjection(nn.Module):
    method __init__ (line 173) | def __init__(self, num_downsampling):
    method forward (line 179) | def forward(self, x):
  class CGNet (line 186) | class CGNet(nn.Module):
    method __init__ (line 215) | def __init__(self,
    method forward (line 309) | def forward(self, x):
    method init_weights (line 338) | def init_weights(self, pretrained=None):
    method train (line 359) | def train(self, mode=True):

FILE: lavis/common/annotator/uniformer/mmseg/models/backbones/fast_scnn.py
  class LearningToDownsample (line 13) | class LearningToDownsample(nn.Module):
    method __init__ (line 29) | def __init__(self,
    method forward (line 66) | def forward(self, x):
  class GlobalFeatureExtractor (line 73) | class GlobalFeatureExtractor(nn.Module):
    method __init__ (line 106) | def __init__(self,
    method _make_layer (line 148) | def _make_layer(self,
    method forward (line 172) | def forward(self, x):
  class FeatureFusionModule (line 181) | class FeatureFusionModule(nn.Module):
    method __init__ (line 199) | def __init__(self,
    method forward (line 235) | def forward(self, higher_res_feature, lower_res_feature):
  class FastSCNN (line 250) | class FastSCNN(nn.Module):
    method __init__ (line 296) | def __init__(self,
    method init_weights (line 360) | def init_weights(self, pretrained=None):
    method forward (line 367) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmseg/models/backbones/hrnet.py
  class HRModule (line 13) | class HRModule(nn.Module):
    method __init__ (line 20) | def __init__(self,
    method _check_branches (line 46) | def _check_branches(self, num_branches, num_blocks, in_channels,
    method _make_one_branch (line 64) | def _make_one_branch(self,
    method _make_branches (line 109) | def _make_branches(self, num_branches, block, num_blocks, num_channels):
    method _make_fuse_layers (line 119) | def _make_fuse_layers(self):
    method forward (line 185) | def forward(self, x):
  class HRNet (line 212) | class HRNet(nn.Module):
    method __init__ (line 273) | def __init__(self,
    method norm1 (line 362) | def norm1(self):
    method norm2 (line 367) | def norm2(self):
    method _make_transition_layer (line 371) | def _make_transition_layer(self, num_channels_pre_layer,
    method _make_layer (line 418) | def _make_layer(self, block, inplanes, planes, blocks, stride=1):
    method _make_stage (line 454) | def _make_stage(self, layer_config, in_channels, multiscale_output=True):
    method init_weights (line 484) | def init_weights(self, pretrained=None):
    method forward (line 510) | def forward(self, x):
    method train (line 547) | def train(self, mode=True):

FILE: lavis/common/annotator/uniformer/mmseg/models/backbones/mobilenet_v2.py
  class MobileNetV2 (line 13) | class MobileNetV2(nn.Module):
    method __init__ (line 45) | def __init__(self,
    method make_layer (line 107) | def make_layer(self, out_channels, num_blocks, stride, dilation,
    method init_weights (line 136) | def init_weights(self, pretrained=None):
    method forward (line 149) | def forward(self, x):
    method _freeze_stages (line 164) | def _freeze_stages(self):
    method train (line 174) | def train(self, mode=True):

FILE: lavis/common/annotator/uniformer/mmseg/models/backbones/mobilenet_v3.py
  class MobileNetV3 (line 15) | class MobileNetV3(nn.Module):
    method __init__ (line 70) | def __init__(self,
    method _make_layer (line 104) | def _make_layer(self):
    method init_weights (line 220) | def init_weights(self, pretrained=None):
    method forward (line 233) | def forward(self, x):
    method _freeze_stages (line 242) | def _freeze_stages(self):
    method train (line 249) | def train(self, mode=True):

FILE: lavis/common/annotator/uniformer/mmseg/models/backbones/resnest.py
  class RSoftmax (line 15) | class RSoftmax(nn.Module):
    method __init__ (line 23) | def __init__(self, radix, groups):
    method forward (line 28) | def forward(self, x):
  class SplitAttentionConv2d (line 39) | class SplitAttentionConv2d(nn.Module):
    method __init__ (line 58) | def __init__(self,
    method norm0 (line 108) | def norm0(self):
    method norm1 (line 113) | def norm1(self):
    method forward (line 117) | def forward(self, x):
  class Bottleneck (line 146) | class Bottleneck(_Bottleneck):
    method __init__ (line 165) | def __init__(self,
    method forward (line 226) | def forward(self, x):
  class ResNeSt (line 270) | class ResNeSt(ResNetV1d):
    method __init__ (line 291) | def __init__(self,
    method make_res_layer (line 305) | def make_res_layer(self, **kwargs):

FILE: lavis/common/annotator/uniformer/mmseg/models/backbones/resnet.py
  class BasicBlock (line 13) | class BasicBlock(nn.Module):
    method __init__ (line 18) | def __init__(self,
    method norm1 (line 58) | def norm1(self):
    method norm2 (line 63) | def norm2(self):
    method forward (line 67) | def forward(self, x):
  class Bottleneck (line 97) | class Bottleneck(nn.Module):
    method __init__ (line 106) | def __init__(self,
    method make_block_plugins (line 219) | def make_block_plugins(self, in_channels, plugins):
    method forward_plugin (line 242) | def forward_plugin(self, x, plugin_names):
    method norm1 (line 250) | def norm1(self):
    method norm2 (line 255) | def norm2(self):
    method norm3 (line 260) | def norm3(self):
    method forward (line 264) | def forward(self, x):
  class ResNet (line 308) | class ResNet(nn.Module):
    method __init__ (line 373) | def __init__(self,
    method make_stage_plugins (line 470) | def make_stage_plugins(self, plugins, stage_idx):
    method make_res_layer (line 523) | def make_res_layer(self, **kwargs):
    method norm1 (line 528) | def norm1(self):
    method _make_stem_layer (line 532) | def _make_stem_layer(self, in_channels, stem_channels):
    method _freeze_stages (line 581) | def _freeze_stages(self):
    method init_weights (line 600) | def init_weights(self, pretrained=None):
    method forward (line 632) | def forward(self, x):
    method train (line 649) | def train(self, mode=True):
  class ResNetV1c (line 662) | class ResNetV1c(ResNet):
    method __init__ (line 672) | def __init__(self, **kwargs):
  class ResNetV1d (line 678) | class ResNetV1d(ResNet):
    method __init__ (line 686) | def __init__(self, **kwargs):

FILE: lavis/common/annotator/uniformer/mmseg/models/backbones/resnext.py
  class Bottleneck (line 11) | class Bottleneck(_Bottleneck):
    method __init__ (line 18) | def __init__(self,
  class ResNeXt (line 87) | class ResNeXt(ResNet):
    method __init__ (line 134) | def __init__(self, groups=1, base_width=4, **kwargs):
    method make_res_layer (line 139) | def make_res_layer(self, **kwargs):

FILE: lavis/common/annotator/uniformer/mmseg/models/backbones/unet.py
  class BasicConvBlock (line 13) | class BasicConvBlock(nn.Module):
    method __init__ (line 43) | def __init__(self,
    method forward (line 76) | def forward(self, x):
  class DeconvModule (line 87) | class DeconvModule(nn.Module):
    method __init__ (line 105) | def __init__(self,
    method forward (line 137) | def forward(self, x):
  class InterpConv (line 148) | class InterpConv(nn.Module):
    method __init__ (line 179) | def __init__(self,
    method forward (line 211) | def forward(self, x):
  class UNet (line 222) | class UNet(nn.Module):
    method __init__ (line 277) | def __init__(self,
    method forward (line 376) | def forward(self, x):
    method train (line 389) | def train(self, mode=True):
    method _check_input_divisible (line 399) | def _check_input_divisible(self, x):
    method init_weights (line 412) | def init_weights(self, pretrained=None):

FILE: lavis/common/annotator/uniformer/mmseg/models/backbones/uniformer.py
  class Mlp (line 24) | class Mlp(nn.Module):
    method __init__ (line 25) | def __init__(self, in_features, hidden_features=None, out_features=Non...
    method forward (line 34) | def forward(self, x):
  class CMlp (line 43) | class CMlp(nn.Module):
    method __init__ (line 44) | def __init__(self, in_features, hidden_features=None, out_features=Non...
    method forward (line 53) | def forward(self, x):
  class CBlock (line 62) | class CBlock(nn.Module):
    method __init__ (line 63) | def __init__(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_sc...
    method forward (line 77) | def forward(self, x):
  class Attention (line 84) | class Attention(nn.Module):
    method __init__ (line 85) | def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, at...
    method forward (line 97) | def forward(self, x):
  class SABlock (line 112) | class SABlock(nn.Module):
    method __init__ (line 113) | def __init__(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_sc...
    method forward (line 128) | def forward(self, x):
  function window_partition (line 138) | def window_partition(x, window_size):
  function window_reverse (line 152) | def window_reverse(windows, window_size, H, W):
  class SABlock_Windows (line 168) | class SABlock_Windows(nn.Module):
    method __init__ (line 169) | def __init__(self, dim, num_heads, window_size=14, mlp_ratio=4., qkv_b...
    method forward (line 185) | def forward(self, x):
  class PatchEmbed (line 218) | class PatchEmbed(nn.Module):
    method __init__ (line 221) | def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=...
    method forward (line 232) | def forward(self, x):
  class UniFormer (line 243) | class UniFormer(nn.Module):
    method __init__ (line 248) | def __init__(self, layers=[3, 4, 8, 3], img_size=224, in_chans=3, num_...
    method init_weights (line 358) | def init_weights(self, pretrained):
    method _init_weights (line 363) | def _init_weights(self, m):
    method no_weight_decay (line 373) | def no_weight_decay(self):
    method get_classifier (line 376) | def get_classifier(self):
    method reset_classifier (line 379) | def reset_classifier(self, num_classes, global_pool=''):
    method forward_features (line 383) | def forward_features(self, x):
    method forward (line 420) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmseg/models/backbones/vit.py
  class Mlp (line 20) | class Mlp(nn.Module):
    method __init__ (line 36) | def __init__(self,
    method forward (line 50) | def forward(self, x):
  class Attention (line 59) | class Attention(nn.Module):
    method __init__ (line 72) | def __init__(self,
    method forward (line 89) | def forward(self, x):
  class Block (line 105) | class Block(nn.Module):
    method __init__ (line 128) | def __init__(self,
    method forward (line 156) | def forward(self, x):
  class PatchEmbed (line 171) | class PatchEmbed(nn.Module):
    method __init__ (line 183) | def __init__(self,
    method forward (line 201) | def forward(self, x):
  class VisionTransformer (line 206) | class VisionTransformer(nn.Module):
    method __init__ (line 246) | def __init__(self,
    method init_weights (line 314) | def init_weights(self, pretrained=None):
    method _pos_embeding (line 359) | def _pos_embeding(self, img, patched_img, pos_embed):
    method resize_pos_embed (line 392) | def resize_pos_embed(pos_embed, input_shpae, pos_shape, patch_size, mo...
    method forward (line 421) | def forward(self, inputs):
    method train (line 454) | def train(self, mode=True):

FILE: lavis/common/annotator/uniformer/mmseg/models/builder.py
  function build_backbone (line 15) | def build_backbone(cfg):
  function build_neck (line 20) | def build_neck(cfg):
  function build_head (line 25) | def build_head(cfg):
  function build_loss (line 30) | def build_loss(cfg):
  function build_segmentor (line 35) | def build_segmentor(cfg, train_cfg=None, test_cfg=None):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/ann_head.py
  class PPMConcat (line 10) | class PPMConcat(nn.ModuleList):
    method __init__ (line 18) | def __init__(self, pool_scales=(1, 3, 6, 8)):
    method forward (line 22) | def forward(self, feats):
  class SelfAttentionBlock (line 32) | class SelfAttentionBlock(_SelfAttentionBlock):
    method __init__ (line 52) | def __init__(self, low_in_channels, high_in_channels, channels,
  class AFNB (line 79) | class AFNB(nn.Module):
    method __init__ (line 99) | def __init__(self, low_in_channels, high_in_channels, channels,
    method forward (line 125) | def forward(self, low_feats, high_feats):
  class APNB (line 133) | class APNB(nn.Module):
    method __init__ (line 150) | def __init__(self, in_channels, channels, out_channels, query_scales,
    method forward (line 175) | def forward(self, feats):
  class ANNHead (line 184) | class ANNHead(BaseDecodeHead):
    method __init__ (line 198) | def __init__(self,
    method forward (line 236) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/apc_head.py
  class ACM (line 11) | class ACM(nn.Module):
    method __init__ (line 25) | def __init__(self, pool_scale, fusion, in_channels, channels, conv_cfg,
    method forward (line 78) | def forward(self, x):
  class APCHead (line 110) | class APCHead(BaseDecodeHead):
    method __init__ (line 124) | def __init__(self, pool_scales=(1, 2, 3, 6), fusion=True, **kwargs):
    method forward (line 149) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/aspp_head.py
  class ASPPModule (line 10) | class ASPPModule(nn.ModuleList):
    method __init__ (line 22) | def __init__(self, dilations, in_channels, channels, conv_cfg, norm_cfg,
    method forward (line 43) | def forward(self, x):
  class ASPPHead (line 53) | class ASPPHead(BaseDecodeHead):
    method __init__ (line 64) | def __init__(self, dilations=(1, 6, 12, 18), **kwargs):
    method forward (line 93) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/cascade_decode_head.py
  class BaseCascadeDecodeHead (line 6) | class BaseCascadeDecodeHead(BaseDecodeHead, metaclass=ABCMeta):
    method __init__ (line 10) | def __init__(self, *args, **kwargs):
    method forward (line 14) | def forward(self, inputs, prev_output):
    method forward_train (line 18) | def forward_train(self, inputs, prev_output, img_metas, gt_semantic_seg,
    method forward_test (line 41) | def forward_test(self, inputs, prev_output, img_metas, test_cfg):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/cc_head.py
  class CCHead (line 13) | class CCHead(FCNHead):
    method __init__ (line 24) | def __init__(self, recurrence=2, **kwargs):
    method forward (line 32) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/da_head.py
  class PAM (line 12) | class PAM(_SelfAttentionBlock):
    method __init__ (line 20) | def __init__(self, in_channels, channels):
    method forward (line 41) | def forward(self, x):
  class CAM (line 49) | class CAM(nn.Module):
    method __init__ (line 52) | def __init__(self):
    method forward (line 56) | def forward(self, x):
  class DAHead (line 75) | class DAHead(BaseDecodeHead):
    method __init__ (line 85) | def __init__(self, pam_channels, **kwargs):
    method pam_cls_seg (line 128) | def pam_cls_seg(self, feat):
    method cam_cls_seg (line 135) | def cam_cls_seg(self, feat):
    method forward (line 142) | def forward(self, inputs):
    method forward_test (line 160) | def forward_test(self, inputs, img_metas, test_cfg):
    method losses (line 164) | def losses(self, seg_logit, seg_label):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/decode_head.py
  class BaseDecodeHead (line 14) | class BaseDecodeHead(nn.Module, metaclass=ABCMeta):
    method __init__ (line 46) | def __init__(self,
    method extra_repr (line 88) | def extra_repr(self):
    method _init_inputs (line 95) | def _init_inputs(self, in_channels, in_index, input_transform):
    method init_weights (line 133) | def init_weights(self):
    method _transform_inputs (line 137) | def _transform_inputs(self, inputs):
    method forward (line 166) | def forward(self, inputs):
    method forward_train (line 170) | def forward_train(self, inputs, img_metas, gt_semantic_seg, train_cfg):
    method forward_test (line 190) | def forward_test(self, inputs, img_metas, test_cfg):
    method cls_seg (line 207) | def cls_seg(self, feat):
    method losses (line 215) | def losses(self, seg_logit, seg_label):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/dm_head.py
  class DCM (line 10) | class DCM(nn.Module):
    method __init__ (line 24) | def __init__(self, filter_size, fusion, in_channels, channels, conv_cfg,
    method forward (line 60) | def forward(self, x):
  class DMHead (line 92) | class DMHead(BaseDecodeHead):
    method __init__ (line 106) | def __init__(self, filter_sizes=(1, 3, 5, 7), fusion=False, **kwargs):
    method forward (line 131) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/dnl_head.py
  class DisentangledNonLocal2d (line 9) | class DisentangledNonLocal2d(NonLocal2d):
    method __init__ (line 16) | def __init__(self, *arg, temperature, **kwargs):
    method embedded_gaussian (line 21) | def embedded_gaussian(self, theta_x, phi_x):
    method forward (line 33) | def forward(self, x):
  class DNLHead (line 87) | class DNLHead(FCNHead):
    method __init__ (line 102) | def __init__(self,
    method forward (line 122) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/ema_head.py
  function reduce_mean (line 13) | def reduce_mean(tensor):
  class EMAModule (line 22) | class EMAModule(nn.Module):
    method __init__ (line 31) | def __init__(self, channels, num_bases, num_stages, momentum):
    method forward (line 44) | def forward(self, feats):
  class EMAHead (line 79) | class EMAHead(BaseDecodeHead):
    method __init__ (line 94) | def __init__(self,
    method forward (line 154) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/enc_head.py
  class EncModule (line 11) | class EncModule(nn.Module):
    method __init__ (line 22) | def __init__(self, in_channels, num_codes, conv_cfg, norm_cfg, act_cfg):
    method forward (line 50) | def forward(self, x):
  class EncHead (line 62) | class EncHead(BaseDecodeHead):
    method __init__ (line 78) | def __init__(self,
    method forward (line 129) | def forward(self, inputs):
    method forward_test (line 151) | def forward_test(self, inputs, img_metas, test_cfg):
    method _convert_to_onehot_labels (line 159) | def _convert_to_onehot_labels(seg_label, num_classes):
    method losses (line 178) | def losses(self, seg_logit, seg_label):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/fcn_head.py
  class FCNHead (line 10) | class FCNHead(BaseDecodeHead):
    method __init__ (line 23) | def __init__(self,
    method forward (line 74) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/fpn_head.py
  class FPNHead (line 11) | class FPNHead(BaseDecodeHead):
    method __init__ (line 23) | def __init__(self, feature_strides, **kwargs):
    method forward (line 54) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/gc_head.py
  class GCHead (line 9) | class GCHead(FCNHead):
    method __init__ (line 23) | def __init__(self,
    method forward (line 38) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/lraspp_head.py
  class LRASPPHead (line 12) | class LRASPPHead(BaseDecodeHead):
    method __init__ (line 23) | def __init__(self, branch_channels=(32, 64), **kwargs):
    method forward (line 68) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/nl_head.py
  class NLHead (line 9) | class NLHead(FCNHead):
    method __init__ (line 23) | def __init__(self,
    method forward (line 40) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/ocr_head.py
  class SpatialGatherModule (line 12) | class SpatialGatherModule(nn.Module):
    method __init__ (line 19) | def __init__(self, scale):
    method forward (line 23) | def forward(self, feats, probs):
  class ObjectAttentionBlock (line 39) | class ObjectAttentionBlock(_SelfAttentionBlock):
    method __init__ (line 42) | def __init__(self, in_channels, channels, scale, conv_cfg, norm_cfg,
    method forward (line 73) | def forward(self, query_feats, key_feats):
  class OCRHead (line 85) | class OCRHead(BaseCascadeDecodeHead):
    method __init__ (line 97) | def __init__(self, ocr_channels, scale=1, **kwargs):
    method forward (line 119) | def forward(self, inputs, prev_output):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/point_head.py
  function calculate_uncertainty (line 14) | def calculate_uncertainty(seg_logits):
  class PointHead (line 35) | class PointHead(BaseCascadeDecodeHead):
    method __init__ (line 60) | def __init__(self,
    method init_weights (line 104) | def init_weights(self):
    method cls_seg (line 108) | def cls_seg(self, feat):
    method forward (line 115) | def forward(self, fine_grained_point_feats, coarse_point_feats):
    method _get_fine_grained_point_feats (line 123) | def _get_fine_grained_point_feats(self, x, points):
    method _get_coarse_point_feats (line 147) | def _get_coarse_point_feats(self, prev_output, points):
    method forward_train (line 165) | def forward_train(self, inputs, prev_output, img_metas, gt_semantic_seg,
    method forward_test (line 203) | def forward_test(self, inputs, prev_output, img_metas, test_cfg):
    method losses (line 248) | def losses(self, point_logits, point_label):
    method get_points_train (line 256) | def get_points_train(self, seg_logits, uncertainty_func, cfg):
    method get_points_test (line 310) | def get_points_test(self, seg_logits, uncertainty_func, cfg):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/psa_head.py
  class PSAHead (line 17) | class PSAHead(BaseDecodeHead):
    method __init__ (line 35) | def __init__(self,
    method forward (line 113) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/psp_head.py
  class PPM (line 10) | class PPM(nn.ModuleList):
    method __init__ (line 24) | def __init__(self, pool_scales, in_channels, channels, conv_cfg, norm_...
    method forward (line 46) | def forward(self, x):
  class PSPHead (line 61) | class PSPHead(BaseDecodeHead):
    method __init__ (line 72) | def __init__(self, pool_scales=(1, 2, 3, 6), **kwargs):
    method forward (line 93) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/sep_aspp_head.py
  class DepthwiseSeparableASPPModule (line 10) | class DepthwiseSeparableASPPModule(ASPPModule):
    method __init__ (line 14) | def __init__(self, **kwargs):
  class DepthwiseSeparableASPPHead (line 29) | class DepthwiseSeparableASPPHead(ASPPHead):
    method __init__ (line 42) | def __init__(self, c1_in_channels, c1_channels, **kwargs):
    method forward (line 78) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/sep_fcn_head.py
  class DepthwiseSeparableFCNHead (line 8) | class DepthwiseSeparableFCNHead(FCNHead):
    method __init__ (line 29) | def __init__(self, **kwargs):

FILE: lavis/common/annotator/uniformer/mmseg/models/decode_heads/uper_head.py
  class UPerHead (line 12) | class UPerHead(BaseDecodeHead):
    method __init__ (line 23) | def __init__(self, pool_scales=(1, 2, 3, 6), **kwargs):
    method psp_forward (line 76) | def psp_forward(self, inputs):
    method forward (line 86) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/losses/accuracy.py
  function accuracy (line 4) | def accuracy(pred, target, topk=1, thresh=None):
  class Accuracy (line 52) | class Accuracy(nn.Module):
    method __init__ (line 55) | def __init__(self, topk=(1, ), thresh=None):
    method forward (line 68) | def forward(self, pred, target):

FILE: lavis/common/annotator/uniformer/mmseg/models/losses/cross_entropy_loss.py
  function cross_entropy (line 9) | def cross_entropy(pred,
  function _expand_onehot_labels (line 35) | def _expand_onehot_labels(labels, label_weights, target_shape, ignore_in...
  function binary_cross_entropy (line 57) | def binary_cross_entropy(pred,
  function mask_cross_entropy (line 100) | def mask_cross_entropy(pred,
  class CrossEntropyLoss (line 139) | class CrossEntropyLoss(nn.Module):
    method __init__ (line 154) | def __init__(self,
    method forward (line 175) | def forward(self,

FILE: lavis/common/annotator/uniformer/mmseg/models/losses/dice_loss.py
  function dice_loss (line 12) | def dice_loss(pred,
  function binary_dice_loss (line 37) | def binary_dice_loss(pred, target, valid_mask, smooth=1, exponent=2, **k...
  class DiceLoss (line 50) | class DiceLoss(nn.Module):
    method __init__ (line 72) | def __init__(self,
    method forward (line 88) | def forward(self,

FILE: lavis/common/annotator/uniformer/mmseg/models/losses/lovasz_loss.py
  function lovasz_grad (line 14) | def lovasz_grad(gt_sorted):
  function flatten_binary_logits (line 29) | def flatten_binary_logits(logits, labels, ignore_index=None):
  function flatten_probs (line 42) | def flatten_probs(probs, labels, ignore_index=None):
  function lovasz_hinge_flat (line 59) | def lovasz_hinge_flat(logits, labels):
  function lovasz_hinge (line 83) | def lovasz_hinge(logits,
  function lovasz_softmax_flat (line 128) | def lovasz_softmax_flat(probs, labels, classes='present', class_weight=N...
  function lovasz_softmax (line 171) | def lovasz_softmax(probs,
  class LovaszLoss (line 225) | class LovaszLoss(nn.Module):
    method __init__ (line 248) | def __init__(self,
    method forward (line 274) | def forward(self,

FILE: lavis/common/annotator/uniformer/mmseg/models/losses/utils.py
  function get_class_weight (line 8) | def get_class_weight(class_weight):
  function reduce_loss (line 26) | def reduce_loss(loss, reduction):
  function weight_reduce_loss (line 46) | def weight_reduce_loss(loss, weight=None, reduction='mean', avg_factor=N...
  function weighted_loss (line 78) | def weighted_loss(loss_func):

FILE: lavis/common/annotator/uniformer/mmseg/models/necks/fpn.py
  class FPN (line 9) | class FPN(nn.Module):
    method __init__ (line 63) | def __init__(self,
    method init_weights (line 157) | def init_weights(self):
    method forward (line 162) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/necks/multilevel_neck.py
  class MultiLevelNeck (line 9) | class MultiLevelNeck(nn.Module):
    method __init__ (line 22) | def __init__(self,
    method forward (line 55) | def forward(self, inputs):

FILE: lavis/common/annotator/uniformer/mmseg/models/segmentors/base.py
  class BaseSegmentor (line 14) | class BaseSegmentor(nn.Module):
    method __init__ (line 19) | def __init__(self):
    method with_neck (line 24) | def with_neck(self):
    method with_auxiliary_head (line 29) | def with_auxiliary_head(self):
    method with_decode_head (line 35) | def with_decode_head(self):
    method extract_feat (line 40) | def extract_feat(self, imgs):
    method encode_decode (line 45) | def encode_decode(self, img, img_metas):
    method forward_train (line 51) | def forward_train(self, imgs, img_metas, **kwargs):
    method simple_test (line 56) | def simple_test(self, img, img_meta, **kwargs):
    method aug_test (line 61) | def aug_test(self, imgs, img_metas, **kwargs):
    method init_weights (line 65) | def init_weights(self, pretrained=None):
    method forward_test (line 76) | def forward_test(self, imgs, img_metas, **kwargs):
    method forward (line 111) | def forward(self, img, img_metas, return_loss=True, **kwargs):
    method train_step (line 126) | def train_step(self, data_batch, optimizer, **kwargs):
    method val_step (line 162) | def val_step(self, data_batch, **kwargs):
    method _parse_losses (line 173) | def _parse_losses(losses):
    method show_result (line 208) | def show_result(self,

FILE: lavis/common/annotator/uniformer/mmseg/models/segmentors/cascade_encoder_decoder.py
  class CascadeEncoderDecoder (line 11) | class CascadeEncoderDecoder(EncoderDecoder):
    method __init__ (line 19) | def __init__(self,
    method _init_decode_head (line 38) | def _init_decode_head(self, decode_head):
    method init_weights (line 48) | def init_weights(self, pretrained=None):
    method encode_decode (line 65) | def encode_decode(self, img, img_metas):
    method _decode_head_forward_train (line 80) | def _decode_head_forward_train(self, x, img_metas, gt_semantic_seg):

FILE: lavis/common/annotator/uniformer/mmseg/models/segmentors/encoder_decoder.py
  class EncoderDecoder (line 13) | class EncoderDecoder(BaseSegmentor):
    method __init__ (line 21) | def __init__(self,
    method _init_decode_head (line 43) | def _init_decode_head(self, decode_head):
    method _init_auxiliary_head (line 49) | def _init_auxiliary_head(self, auxiliary_head):
    method init_weights (line 59) | def init_weights(self, pretrained=None):
    method extract_feat (line 77) | def extract_feat(self, img):
    method encode_decode (line 84) | def encode_decode(self, img, img_metas):
    method _decode_head_forward_train (line 96) | def _decode_head_forward_train(self, x, img_metas, gt_semantic_seg):
    method _decode_head_forward_test (line 107) | def _decode_head_forward_test(self, x, img_metas):
    method _auxiliary_head_forward_train (line 113) | def _auxiliary_head_forward_train(self, x, img_metas, gt_semantic_seg):
    method forward_dummy (line 130) | def forward_dummy(self, img):
    method forward_train (line 136) | def forward_train(self, img, img_metas, gt_semantic_seg):
    method slide_inference (line 169) | def slide_inference(self, img, img_meta, rescale):
    method whole_inference (line 214) | def whole_inference(self, img, img_meta, rescale):
    method inference (line 233) | def inference(self, img, img_meta, rescale):
    method simple_test (line 268) | def simple_test(self, img, img_meta, rescale=True):
    method aug_test (line 281) | def aug_test(self, imgs, img_metas, rescale=True):

FILE: lavis/common/annotator/uniformer/mmseg/models/utils/drop.py
  class DropPath (line 8) | class DropPath(nn.Module):
    method __init__ (line 17) | def __init__(self, drop_prob=0.):
    method forward (line 22) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmseg/models/utils/inverted_residual.py
  class InvertedResidual (line 8) | class InvertedResidual(nn.Module):
    method __init__ (line 31) | def __init__(self,
    method forward (line 81) | def forward(self, x):
  class InvertedResidualV3 (line 97) | class InvertedResidualV3(nn.Module):
    method __init__ (line 124) | def __init__(self,
    method forward (line 183) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmseg/models/utils/make_divisible.py
  function make_divisible (line 1) | def make_divisible(value, divisor, min_value=None, min_ratio=0.9):

FILE: lavis/common/annotator/uniformer/mmseg/models/utils/res_layer.py
  class ResLayer (line 5) | class ResLayer(nn.Sequential):
    method __init__ (line 26) | def __init__(self,

FILE: lavis/common/annotator/uniformer/mmseg/models/utils/se_layer.py
  class SELayer (line 8) | class SELayer(nn.Module):
    method __init__ (line 26) | def __init__(self,
    method forward (line 53) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmseg/models/utils/self_attention_block.py
  class SelfAttentionBlock (line 7) | class SelfAttentionBlock(nn.Module):
    method __init__ (line 32) | def __init__(self, key_in_channels, query_in_channels, channels,
    method init_weights (line 93) | def init_weights(self):
    method build_project (line 99) | def build_project(self, in_channels, channels, num_convs, use_conv_mod...
    method forward (line 131) | def forward(self, query_feats, key_feats):

FILE: lavis/common/annotator/uniformer/mmseg/models/utils/up_conv_block.py
  class UpConvBlock (line 6) | class UpConvBlock(nn.Module):
    method __init__ (line 44) | def __init__(self,
    method forward (line 94) | def forward(self, skip, x):

FILE: lavis/common/annotator/uniformer/mmseg/models/utils/weight_init.py
  function _no_grad_trunc_normal_ (line 10) | def _no_grad_trunc_normal_(tensor, mean, std, a, b):
  function trunc_normal_ (line 48) | def trunc_normal_(tensor, mean=0., std=1., a=-2., b=2.):

FILE: lavis/common/annotator/uniformer/mmseg/ops/encoding.py
  class Encoding (line 6) | class Encoding(nn.Module):
    method __init__ (line 17) | def __init__(self, channels, num_codes):
    method scaled_l2 (line 33) | def scaled_l2(x, codewords, scale):
    method aggregate (line 46) | def aggregate(assignment_weights, x, codewords):
    method forward (line 57) | def forward(self, x):
    method __repr__ (line 70) | def __repr__(self):

FILE: lavis/common/annotator/uniformer/mmseg/ops/wrappers.py
  function resize (line 7) | def resize(input,
  class Upsample (line 29) | class Upsample(nn.Module):
    method __init__ (line 31) | def __init__(self,
    method forward (line 45) | def forward(self, x):

FILE: lavis/common/annotator/uniformer/mmseg/utils/collect_env.py
  function collect_env (line 7) | def collect_env():

FILE: lavis/common/annotator/uniformer/mmseg/utils/logger.py
  function get_root_logger (line 6) | def get_root_logger(log_file=None, log_level=logging.INFO):

FILE: lavis/common/annotator/util.py
  function HWC3 (line 9) | def HWC3(x):
  function resize_image (line 28) | def resize_image(input_image, resolution):

FILE: lavis/common/config.py
  class Config (line 16) | class Config:
    method __init__ (line 17) | def __init__(self, args):
    method _validate_runner_config (line 43) | def _validate_runner_config(self, runner_config):
    method _build_opt_list (line 52) | def _build_opt_list(self, opts):
    method build_model_config (line 57) | def build_model_config(config, **kwargs):
    method build_runner_config (line 84) | def build_runner_config(config):
    method build_dataset_config (line 88) | def build_dataset_config(config):
    method _convert_to_dot_list (line 114) | def _convert_to_dot_list(self, opts):
    method get_config (line 128) | def get_config(self):
    method run_cfg (line 132) | def run_cfg(self):
    method datasets_cfg (line 136) | def datasets_cfg(self):
    method model_cfg (line 140) | def model_cfg(self):
    method pretty_print (line 143) | def pretty_print(self):
    method _convert_node_to_json (line 161) | def _convert_node_to_json(self, node):
    method to_dict (line 165) | def to_dict(self):
  function node_to_dict (line 169) | def node_to_dict(node):
  class ConfigValidator (line 173) | class ConfigValidator:
    class _Argument (line 187) | class _Argument:
      method __init__ (line 188) | def __init__(self, name, choices=None, type=None, help=None):
      method __str__ (line 195) | def __str__(self):
    method __init__ (line 205) | def __init__(self, description):
    method __getitem__ (line 212) | def __getitem__(self, key):
    method __str__ (line 217) | def __str__(self) -> str:
    method add_argument (line 220) | def add_argument(self, *args, **kwargs):
    method validate (line 226) | def validate(self, config=None):
    method format_arguments (line 248) | def format_arguments(self):
    method format_help (line 251) | def format_help(self):
    method print_help (line 256) | def print_help(self):
  function create_runner_config_validator (line 261) | def create_runner_config_validator():

FILE: lavis/common/dist_utils.py
  function setup_for_distributed (line 17) | def setup_for_distributed(is_master):
  function is_dist_avail_and_initialized (line 33) | def is_dist_avail_and_initialized():
  function get_world_size (line 41) | def get_world_size():
  function get_rank (line 47) | def get_rank():
  function is_main_process (line 53) | def is_main_process():
  function init_distributed_mode (line 57) | def init_distributed_mode(args):
  function get_dist_info (line 93) | def get_dist_info():
  function main_process (line 107) | def main_process(func):
  function download_cached_file (line 117) | def download_cached_file(url, check_hash=True, progress=False):

FILE: lavis/common/gradcam.py
  function getAttMap (line 7) | def getAttMap(img, attMap, blur=True, overlap=True):

FILE: lavis/common/logger.py
  class SmoothedValue (line 19) | class SmoothedValue(object):
    method __init__ (line 24) | def __init__(self, window_size=20, fmt=None):
    method update (line 32) | def update(self, value, n=1):
    method synchronize_between_processes (line 37) | def synchronize_between_processes(self):
    method median (line 51) | def median(self):
    method avg (line 56) | def avg(self):
    method global_avg (line 61) | def global_avg(self):
    method max (line 65) | def max(self):
    method value (line 69) | def value(self):
    method __str__ (line 72) | def __str__(self):
  class MetricLogger (line 82) | class MetricLogger(object):
    method __init__ (line 83) | def __init__(self, delimiter="\t"):
    method update (line 87) | def update(self, **kwargs):
    method __getattr__ (line 94) | def __getattr__(self, attr):
    method __str__ (line 103) | def __str__(self):
    method global_avg (line 109) | def global_avg(self):
    method synchronize_between_processes (line 115) | def synchronize_between_processes(self):
    method add_meter (line 119) | def add_meter(self, name, meter):
    method log_every (line 122) | def log_every(self, iterable, print_freq, header=None):
  class AttrDict (line 184) | class AttrDict(dict):
    method __init__ (line 185) | def __init__(self, *args, **kwargs):
  function setup_logger (line 190) | def setup_logger():

FILE: lavis/common/optims.py
  class LinearWarmupStepLRScheduler (line 14) | class LinearWarmupStepLRScheduler:
    method __init__ (line 15) | def __init__(
    method step (line 37) | def step(self, cur_epoch, cur_step):
  class LinearWarmupCosineLRScheduler (line 57) | class LinearWarmupCosineLRScheduler:
    method __init__ (line 58) | def __init__(
    method step (line 77) | def step(self, cur_epoch, cur_step):
  class ConstantLRScheduler (line 98) | class ConstantLRScheduler:
    method __init__ (line 99) | def __init__(self, optimizer, init_lr, warmup_start_lr=-1, warmup_step...
    method step (line 105) | def step(self, cur_epoch, cur_step):
  function cosine_lr_schedule (line 119) | def cosine_lr_schedule(optimizer, epoch, max_epoch, init_lr, min_lr):
  function warmup_lr_schedule (line 128) | def warmup_lr_schedule(optimizer, step, max_step, init_lr, max_lr):
  function step_lr_schedule (line 135) | def step_lr_schedule(optimizer, epoch, init_lr, min_lr, decay_rate):

FILE: lavis/common/registry.py
  class Registry (line 9) | class Registry:
    method register_builder (line 22) | def register_builder(cls, name):
    method register_task (line 54) | def register_task(cls, name):
    method register_model (line 83) | def register_model(cls, name):
    method register_processor (line 112) | def register_processor(cls, name):
    method register_lr_scheduler (line 141) | def register_lr_scheduler(cls, name):
    method register_runner (line 165) | def register_runner(cls, name):
    method register_path (line 189) | def register_path(cls, name, path):
    method register (line 205) | def register(cls, name, obj):
    method get_builder_class (line 232) | def get_builder_class(cls, name):
    method get_model_class (line 236) | def get_model_class(cls, name):
    method get_task_class (line 240) | def get_task_class(cls, name):
    method get_processor_class (line 244) | def get_processor_class(cls, name):
    method get_lr_scheduler_class (line 248) | def get_lr_scheduler_class(cls, name):
    method get_runner_class (line 252) | def get_runner_class(cls, name):
    method list_runners (line 256) | def list_runners(cls):
    method list_models (line 260) | def list_models(cls):
    method list_tasks (line 264) | def list_tasks(cls):
    method list_processors (line 268) | def list_processors(cls):
    method list_lr_schedulers (line 272) | def list_lr_schedulers(cls):
    method list_datasets (line 276) | def list_datasets(cls):
    method get_path (line 280) | def get_path(cls, name):
    method get (line 284) | def get(cls, name, default=None, no_warning=False):
    method unregister (line 315) | def unregister(cls, name):

FILE: lavis/common/utils.py
  function now (line 37) | def now():
  function is_url (line 43) | def is_url(url_or_filename):
  function get_cache_path (line 48) | def get_cache_path(rel_path):
  function get_abs_path (line 52) | def get_abs_path(rel_path):
  function load_json (line 56) | def load_json(filename):
  function makedir (line 66) | def makedir(dir_path):
  function get_redirected_url (line 80) | def get_redirected_url(url: str):
  function to_google_drive_download_url (line 95) | def to_google_drive_download_url(view_url: str) -> str:
  function download_google_drive_url (line 110) | def download_google_drive_url(url: str, output_path: str, output_file_na...
  function _get_google_drive_file_id (line 143) | def _get_google_drive_file_id(url: str) -> Optional[str]:
  function _urlretrieve (line 156) | def _urlretrieve(url: str, filename: str, chunk_size: int = 1024) -> None:
  function download_url (line 169) | def download_url(
  function download_and_extract_archive (line 223) | def download_and_extract_archive(
  function cache_url (line 244) | def cache_url(url: str, cache_dir: str) -> str:
  function create_file_symlink (line 263) | def create_file_symlink(file1, file2):
  function save_file (line 277) | def save_file(data, filename, append_to_json=True, verbose=True):
  function load_file (line 315) | def load_file(filename, mmap_mode=None, verbose=True, allow_pickle=False):
  function abspath (line 376) | def abspath(resource_path: str):
  function makedir (line 388) | def makedir(dir_path):
  function is_url (line 402) | def is_url(input_url):
  function download_and_untar (line 410) | def download_and_untar(url):
  function cleanup_dir (line 426) | def cleanup_dir(dir):
  function get_file_size (line 437) | def get_file_size(filename):
  function is_serializable (line 444) | def is_serializable(value):
  function is_convertible_to_int (line 454) | def is_convertible_to_int(value):

FILE: lavis/common/vqa_tools/vqa.py
  class VQA (line 31) | class VQA:
    method __init__ (line 32) | def __init__(self, annotation_file=None, question_file=None):
    method createIndex (line 53) | def createIndex(self):
    method info (line 71) | def info(self):
    method getQuesIds (line 79) | def getQuesIds(self, imgIds=[], quesTypes=[], ansTypes=[]):
    method getImgIds (line 114) | def getImgIds(self, quesIds=[], quesTypes=[], ansTypes=[]):
    method loadQA (line 148) | def loadQA(self, ids=[]):
    method showQA (line 159) | def showQA(self, anns):
    method loadRes (line 173) | def loadRes(self, resFile, quesFile):

FILE: lavis/common/vqa_tools/vqa_eval.py
  class VQAEval (line 18) | class VQAEval:
    method __init__ (line 19) | def __init__(self, vqa=None, vqaRes=None, n=2):
    method evaluate (line 193) | def evaluate(self, quesIds=None):
    method processPunctuation (line 249) | def processPunctuation(self, inText):
    method processDigitArticle (line 261) | def processDigitArticle(self, inText):
    method setAccuracy (line 276) | def setAccuracy(self, accQA, accQuesType, accAnsType):
    method setEvalQA (line 292) | def setEvalQA(self, quesId, acc):
    method setEvalQuesType (line 295) | def setEvalQuesType(self, quesId, quesType, acc):
    method setEvalAnsType (line 300) | def setEvalAnsType(self, quesId, ansType, acc):
    method updateProgress (line 305) | def updateProgress(self, progress):

FILE: lavis/datasets/builders/__init__.py
  function load_dataset (line 230) | def load_dataset(name, cfg_path=None, vis_path=None, data_type=None):
  class DatasetZoo (line 268) | class DatasetZoo:
    method __init__ (line 269) | def __init__(self) -> None:
    method get_names (line 275) | def get_names(self):

FILE: lavis/datasets/builders/audio_caption_builder.py
  class AudioCapBuilder (line 27) | class AudioCapBuilder(MultiModalDatasetBuilder):
    method build (line 36) | def build(self):
  class AudioSetBuilder (line 49) | class AudioSetBuilder(AudioCapBuilder):
  class AudioSetInstructBuilder (line 58) | class AudioSetInstructBuilder(AudioCapBuilder):
  class AudioCapsCapBuilder (line 67) | class AudioCapsCapBuilder(AudioCapBuilder):
  class AudioCapsInstructCapBuilder (line 76) | class AudioCapsInstructCapBuilder(AudioCapBuilder):
  class ClothoCapInstructBuilder (line 85) | class ClothoCapInstructBuilder(MultiModalDatasetBuilder):
  class ClothoCapInstructBuilder (line 94) | class ClothoCapInstructBuilder(MultiModalDatasetBuilder):
  class WavCapsCapBuilder (line 104) | class WavCapsCapBuilder(AudioCapBuilder):
  class WavCapsCapInstructBuilder (line 115) | class WavCapsCapInstructBuilder(AudioCapBuilder):

FILE: lavis/datasets/builders/audio_qa_builder.py
  class AudioCapsQABuilder (line 13) | class AudioCapsQABuilder(AudioCapBuilder):
  class ClothoQABuilder (line 22) | class ClothoQABuilder(AudioCapBuilder):

FILE: lavis/datasets/builders/base_dataset_builder.py
  class BaseDatasetBuilder (line 23) | class BaseDatasetBuilder:
    method __init__ (line 26) | def __init__(self, cfg=None):
    method build_datasets (line 46) | def build_datasets(self):
    method build_processors (line 62) | def build_processors(self):
    method _build_proc_from_cfg (line 86) | def _build_proc_from_cfg(cfg):
    method default_config_path (line 94) | def default_config_path(cls, type="default"):
    method _download_data (line 97) | def _download_data(self):
    method _download_ann (line 101) | def _download_ann(self):
    method _download_vis (line 158) | def _download_vis(self):
    method build (line 172) | def build(self):
  class MultiModalDatasetBuilder (line 238) | class MultiModalDatasetBuilder(BaseDatasetBuilder):
    method __init__ (line 247) | def __init__(self, cfg=None):
    method _build_processor (line 252) | def _build_processor(self, cfg_name):
    method build_processors (line 261) | def build_processors(self):
    method _download_multimodal (line 274) | def _download_multimodal(self, modality):
    method _download_data (line 279) | def _download_data(self):
    method _get_absolute_path (line 284) | def _get_absolute_path(self, path):
    method build (line 289) | def build(self):
    method _get_dataset_args (line 306) | def _get_dataset_args(self, info, is_train):
  function load_dataset_config (line 325) | def load_dataset_config(cfg_path):

FILE: lavis/datasets/builders/caption_builder.py
  class COCOCapBuilder (line 40) | class COCOCapBuilder(BaseDatasetBuilder):
  class COCOCapInstructBuilder (line 49) | class COCOCapInstructBuilder(BaseDatasetBuilder):
  class Flickr30kCapBuilder (line 59) | class Flickr30kCapBuilder(BaseDatasetBuilder):
  class Flickr30kCapInstructBuilder (line 67) | class Flickr30kCapInstructBuilder(BaseDatasetBuilder):
  class COCOCapBuilder (line 75) | class COCOCapBuilder(BaseDatasetBuilder):
  class VSRCapBuilder (line 83) | class VSRCapBuilder(BaseDatasetBuilder):
  class VSRCapInstructBuilder (line 92) | class VSRCapInstructBuilder(BaseDatasetBuilder):
  class TextCapsCapBuilder (line 101) | class TextCapsCapBuilder(BaseDatasetBuilder):
  class TextCapsCapInstructBuilder (line 110) | class TextCapsCapInstructBuilder(BaseDatasetBuilder):
  class CapFiltCapBuilder (line 120) | class CapFiltCapBuilder(BaseDatasetBuilder):
  class CapFiltCapBuilder (line 128) | class CapFiltCapBuilder(BaseDatasetBuilder):
  class MSRVTTCapBuilder (line 137) | class MSRVTTCapBuilder(BaseDatasetBuilder):
  class MSVDCapBuilder (line 147) | class MSVDCapBuilder(BaseDatasetBuilder):
  class VATEXCapBuilder (line 157) | class VATEXCapBuilder(MultiModalDatasetBuilder):
  class MSRVTTCapInstructBuilder (line 166) | class MSRVTTCapInstructBuilder(BaseDatasetBuilder):
  class MSVDCapInstructBuilder (line 175) | class MSVDCapInstructBuilder(BaseDatasetBuilder):
  class VATEXCapInstructBuilder (line 186) | class VATEXCapInstructBuilder(MultiModalDatasetBuilder):
  class WebVid2MCapBuilder (line 196) | class WebVid2MCapBuilder(BaseDatasetBuilder):
  class WebVid2MCapInstructBuilder (line 204) | class WebVid2MCapInstructBuilder(BaseDatasetBuilder):
  class ViolinCapBuilder (line 212) | class ViolinCapBuilder(BaseDatasetBuilder):
  class ViolinCapInstructBuilder (line 222) | class ViolinCapInstructBuilder(BaseDatasetBuilder):
  class VALORCaptionBuilder (line 231) | class VALORCaptionBuilder(MultiModalDatasetBuilder):
  class VALORCaptionInstructBuilder (line 240) | class VALORCaptionInstructBuilder(MultiModalDatasetBuilder):
  class VlepCaptionBuilder (line 249) | class VlepCaptionBuilder(BaseDatasetBuilder):
  class VlepCaptionInstructBuilder (line 259) | class VlepCaptionInstructBuilder(BaseDatasetBuilder):
  class YouCookCaptionBuilder (line 268) | class YouCookCaptionBuilder(BaseDatasetBuilder):
  class YouCookCaptionInstructBuilder (line 277) | class YouCookCaptionInstructBuilder(BaseDatasetBuilder):
  class COINCaptionBuilder (line 286) | class COINCaptionBuilder(BaseDatasetBuilder):
  class COINCaptionInstructBuilder (line 296) | class COINCaptionInstructBuilder(BaseDatasetBuilder):
  class CharadeCaptionBuilder (line 306) | class CharadeCaptionBuilder(BaseDatasetBuilder):
  class CharadeCaptionInstructBuilder (line 315) | class CharadeCaptionInstructBuilder(BaseDatasetBuilder):

FILE: lavis/datasets/builders/classification_builder.py
  class ViolinEntailmentBuilder (line 16) | class ViolinEntailmentBuilder(BaseDatasetBuilder):
  class ViolinEntailmentInstructBuilder (line 26) | class ViolinEntailmentInstructBuilder(BaseDatasetBuilder):
  class NLVRBuilder (line 35) | class NLVRBuilder(BaseDatasetBuilder):
  class SNLIVisualEntailmentBuilder (line 43) | class SNLIVisualEntailmentBuilder(BaseDatasetBuilder):
  class SNLIVisualEntailmentInstructBuilder (line 50) | class SNLIVisualEntailmentInstructBuilder(BaseDatasetBuilder):
  class VSRClassificationBuilder (line 58) | class VSRClassificationBuilder(BaseDatasetBuilder):
  class SNLIVisualEntailmentInstructBuilder (line 65) | class SNLIVisualEntailmentInstructBuilder(BaseDatasetBuilder):
  class ESC50ClassificationBuilder (line 72) | class ESC50ClassificationBuilder(MultiModalDatasetBuilder):

FILE: lavis/datasets/builders/dialogue_builder.py
  class AVSDDialBuilder (line 26) | class AVSDDialBuilder(BaseDatasetBuilder):
  class VisDialBuilder (line 33) | class VisDialBuilder(BaseDatasetBuilder):
  class VisDialInstructBuilder (line 40) | class VisDialInstructBuilder(BaseDatasetBuilder):
  class AVSDDialInstructBuilder (line 47) | class AVSDDialInstructBuilder(MultiModalDatasetBuilder):
  class LLaVA150kDialInstructBuilder (line 54) | class LLaVA150kDialInstructBuilder(BaseDatasetBuilder):
  class YT8MDialBuilder (line 61) | class YT8MDialBuilder(MultiModalDatasetBuilder):

FILE: lavis/datasets/builders/discrn_builders.py
  class DiscrnImagePcBuilder (line 15) | class DiscrnImagePcBuilder(MultiModalDatasetBuilder):
  class DiscrnAudioVideoBuilder (line 23) | class DiscrnAudioVideoBuilder(MultiModalDatasetBuilder):

FILE: lavis/datasets/builders/image_text_pair_builder.py
  class ConceptualCaption3MBuilder (line 16) | class ConceptualCaption3MBuilder(BaseDatasetBuilder):
  class ConceptualCaption3MInstructBuilder (line 24) | class ConceptualCaption3MInstructBuilder(BaseDatasetBuilder):
  class ConceptualCaption12MBuilder (line 33) | class ConceptualCaption12MBuilder(BaseDatasetBuilder):
  class ConceptualCaption12MInstructBuilder (line 41) | class ConceptualCaption12MInstructBuilder(BaseDatasetBuilder):
  class SBUCaptionBuilder (line 49) | class SBUCaptionBuilder(BaseDatasetBuilder):
  class SBUCaptionInstructBuilder (line 56) | class SBUCaptionInstructBuilder(BaseDatasetBuilder):
  class VGCaptionBuilder (line 63) | class VGCaptionBuilder(BaseDatasetBuilder):
  class VGCaptionInstructBuilder (line 70) | class VGCaptionInstructBuilder(BaseDatasetBuilder):
  class Laion2BMultiBuilder (line 78) | class Laion2BMultiBuilder(BaseDatasetBuilder):
    method _download_ann (line 83) | def _download_ann(self):
    method _download_vis (line 86) | def _download_vis(self):
    method build (line 89) | def build(self):
  class Laion400MBuilder (line 109) | class Laion400MBuilder(Laion2BMultiBuilder):
  class Laion400MInstructBuilder (line 116) | class Laion400MInstructBuilder(Laion2BMultiBuilder):

FILE: lavis/datasets/builders/imagefolder_builder.py
  class ImageNetBuilder (line 16) | class ImageNetBuilder(BaseDatasetBuilder):
    method _download_ann (line 22) | def _download_ann(self):
    method build (line 25) | def build(self):

FILE: lavis/datasets/builders/object3d_caption_builder.py
  class ObjaverseCaptionBuilder (line 20) | class ObjaverseCaptionBuilder(MultiModalDatasetBuilder):
    method build (line 28) | def build(self):
  class ObjaverseCaptionInstructBuilder (line 41) | class ObjaverseCaptionInstructBuilder(ObjaverseCaptionBuilder):
  class ShapenetCaptionBuilder (line 50) | class ShapenetCaptionBuilder(ObjaverseCaptionBuilder):
  class ShapenetCaptionInstructBuilder (line 59) | class ShapenetCaptionInstructBuilder(ObjaverseCaptionBuilder):

FILE: lavis/datasets/builders/object3d_classification_builder.py
  class ModelNetClassificationBuilder (line 13) | class ModelNetClassificationBuilder(MultiModalDatasetBuilder):

FILE: lavis/datasets/builders/object3d_qa_builder.py
  class ObjaverseQABuilder (line 13) | class ObjaverseQABuilder(ObjaverseCaptionBuilder):

FILE: lavis/datasets/builders/retrieval_builder.py
  class MSRVTTRetrievalBuilder (line 20) | class MSRVTTRetrievalBuilder(BaseDatasetBuilder):
  class DiDeMoRetrievalBuilder (line 28) | class DiDeMoRetrievalBuilder(BaseDatasetBuilder):
  class COCORetrievalBuilder (line 36) | class COCORetrievalBuilder(BaseDatasetBuilder):
  class Flickr30kBuilder (line 44) | class Flickr30kBuilder(BaseDatasetBuilder):

FILE: lavis/datasets/builders/text_to_image_generation_builder.py
  class BlipDiffusionFinetuneBuilder (line 16) | class BlipDiffusionFinetuneBuilder(BaseDatasetBuilder):
    method _download_ann (line 23) | def _download_ann(self):
    method build (line 26) | def build(self):

FILE: lavis/datasets/builders/video_qa_builder.py
  class VideoQABuilder (line 15) | class VideoQABuilder(BaseDatasetBuilder):
    method build (line 19) | def build(self):
  class MSRVTTQABuilder (line 35) | class MSRVTTQABuilder(VideoQABuilder):
  class MSVDQABuilder (line 42) | class MSVDQABuilder(VideoQABuilder):
  class MSRVTTQAInstructBuilder (line 49) | class MSRVTTQAInstructBuilder(VideoQABuilder):
  class MSVDQAInstructBuilder (line 58) | class MSVDQAInstructBuilder(VideoQABuilder):
  class MusicAVQABuilder (line 66) | class MusicAVQABuilder(MultiModalDatasetBuilder):
  class MusicAVQAInstructBuilder (line 73) | class MusicAVQAInstructBuilder(MultiModalDatasetBuilder):

FILE: lavis/datasets/builders/vqa_builder.py
  class COCOVQABuilder (line 20) | class COCOVQABuilder(BaseDatasetBuilder):
  class COCOVQAInstructBuilder (line 30) | class COCOVQAInstructBuilder(BaseDatasetBuilder):
  class VGVQABuilder (line 40) | class VGVQABuilder(BaseDatasetBuilder):
  class VGVQAInstructBuilder (line 45) | class VGVQAInstructBuilder(BaseDatasetBuilder):
  class OKVQABuilder (line 50) | class OKVQABuilder(COCOVQABuilder):
  class OKVQAInstructBuilder (line 56) | class OKVQAInstructBuilder(COCOVQAInstructBuilder):
  class AOKVQABuilder (line 62) | class AOKVQABuilder(BaseDatasetBuilder):
  class AOKVQAInstructBuilder (line 69) | class AOKVQAInstructBuilder(BaseDatasetBuilder):
  class GQABuilder (line 77) | class GQABuilder(BaseDatasetBuilder):
  class GQAInstructBuilder (line 88) | class GQAInstructBuilder(BaseDatasetBuilder):
  class IconQABuilder (line 99) | class IconQABuilder(BaseDatasetBuilder):
  class IconQAInstructBuilder (line 108) | class IconQAInstructBuilder(BaseDatasetBuilder):
  class ScienceQABuilder (line 117) | class ScienceQABuilder(BaseDatasetBuilder):
  class ScienceQAInstructBuilder (line 124) | class ScienceQAInstructBuilder(BaseDatasetBuilder):
  class OCRVQABuilder (line 131) | class OCRVQABuilder(BaseDatasetBuilder):
  class OCRVQAInstructBuilder (line 138) | class OCRVQAInstructBuilder(BaseDatasetBuilder):
  class VizWizVQABuilder (line 146) | class VizWizVQABuilder(BaseDatasetBuilder):

FILE: lavis/datasets/data_utils.py
  function load_video (line 30) | def load_video(video_path, n_frms=MAX_INT, height=-1, width=-1, sampling...
  function apply_to_sample (line 53) | def apply_to_sample(f, sample):
  function move_to_cuda (line 71) | def move_to_cuda(sample):
  function prepare_sample (line 78) | def prepare_sample(samples, cuda_enabled=True):
  function reorg_datasets_by_split (line 87) | def reorg_datasets_by_split(datasets):
  function concat_datasets (line 113) | def concat_datasets(datasets):
  function extract_archive (line 180) | def extract_archive(from_path, to_path=None, overwrite=False):
  function save_frames_grid (line 262) | def save_frames_grid(img_array, out_path):
  function uniform_frame_sampling (line 289) | def uniform_frame_sampling(video_path, num_frames, target_height, target...
  function head_tail_frame_sampling (line 316) | def head_tail_frame_sampling(video_path, num_frames, target_height, targ...
  function load_clip (line 345) | def load_clip(video_path, num_frames, target_height, target_width, start...

FILE: lavis/datasets/datasets/aok_vqa_datasets.py
  class __DisplMixin (line 19) | class __DisplMixin:
    method displ_item (line 20) | def displ_item(self, index):
  class AOKVQADataset (line 35) | class AOKVQADataset(VQADataset, __DisplMixin):
    method __init__ (line 36) | def __init__(self, vis_processor, text_processor, vis_root, ann_paths):
    method __getitem__ (line 39) | def __getitem__(self, index):
  class AOKVQAInstructDataset (line 67) | class AOKVQAInstructDataset(AOKVQADataset):
    method __getitem__ (line 68) | def __getitem__(self, index):
    method collater (line 74) | def collater(self, samples):
  class AOKVQAEvalDataset (line 80) | class AOKVQAEvalDataset(VQAEvalDataset, __DisplMixin):
    method __init__ (line 81) | def __init__(self, vis_processor, text_processor, vis_root, ann_paths):
    method collater (line 109) | def collater(self, samples):
    method __getitem__ (line 139) | def __getitem__(self, index):

FILE: lavis/datasets/datasets/audio_captioning_datasets.py
  class __DisplMixin (line 22) | class __DisplMixin:
    method displ_item (line 23) | def displ_item(self, index):
  class AudioCaptioningDataset (line 38) | class AudioCaptioningDataset(BaseDataset, __DisplMixin):
    method __init__ (line 39) | def __init__(self, **kwargs):
    method get_audio_path (line 47) | def get_audio_path(self, ann):
    method is_empty_audio (line 50) | def is_empty_audio(self, ann):
    method get_existing_audio_annotations (line 64) | def get_existing_audio_annotations(self):
    method get_existing_video_annotations (line 67) | def get_existing_video_annotations(self):
    method get_existing_images_annotations (line 70) | def get_existing_images_annotations(self):
    method get_video_path (line 73) | def get_video_path(self, ann):
    method get_images_path (line 76) | def get_images_path(self, ann):
    method __len__ (line 79) | def __len__(self):
    method __getitem__ (line 82) | def __getitem__(self, index):
    method _build_templates (line 85) | def _build_templates(self, templates_path):
  class AudioSetDataset (line 93) | class AudioSetDataset(AudioCaptioningDataset):
    method __init__ (line 94) | def __init__(self, **kwargs):
    method get_audio_path (line 113) | def get_audio_path(self, ann):
    method __getitem__ (line 121) | def __getitem__(self, index):
  class AudioSetInstructDataset (line 149) | class AudioSetInstructDataset(AudioSetDataset):
    method __getitem__ (line 150) | def __getitem__(self, index):
  class AudioSetEvalDataset (line 157) | class AudioSetEvalDataset(AudioSetDataset):
    method __getitem__ (line 158) | def __getitem__(self, index):
  class AudioCapsDataset (line 164) | class AudioCapsDataset(AudioCaptioningDataset):
    method __init__ (line 165) | def __init__(self, **kwargs):
    method get_audio_path (line 182) | def get_audio_path(self, ann):
    method get_cached_audio_path (line 188) | def get_cached_audio_path(self, ann):
    method __getitem__ (line 194) | def __getitem__(self, index):
  class AudioCapsInstructDataset (line 219) | class AudioCapsInstructDataset(AudioCapsDataset):
    method __getitem__ (line 220) | def __getitem__(self, index):
  class AudioCapsEvalDataset (line 227) | class AudioCapsEvalDataset(AudioCapsDataset):
    method __init__ (line 228) | def __init__(self, **kwargs):
    method __getitem__ (line 233) | def __getitem__(self, index):
  class ClothoV2Dataset (line 239) | class ClothoV2Dataset(BaseDataset, __DisplMixin):
    method __init__ (line 240) | def __init__(self, **kwargs):
    method __getitem__ (line 260) | def __getitem__(self, index):
  class ClothoV2InstructDataset (line 269) | class ClothoV2InstructDataset(ClothoV2Dataset):
    method __getitem__ (line 270) | def __getitem__(self, index):
  class ClothoV2EvalDataset (line 277) | class ClothoV2EvalDataset(ClothoV2Dataset):
    method __getitem__ (line 278) | def __getitem__(self, index):
  class AudioLanguagePretrainDataset (line 313) | class AudioLanguagePretrainDataset(BaseDataset, __DisplMixin):
    method __init__ (line 314) | def __init__(self, **kwargs):
    method _load_json_file (line 326) | def _load_json_file(self, files, audio_root, blacklist=None):
    method __len__ (line 366) | def __len__(self):
    method __getitem__ (line 369) | def __getitem__(self, index):
    method _build_templates (line 391) | def _build_templates(self, templates_path)

[
  {
    "path": ".github/workflows/docs.yaml",
    "chars": 888,
    "preview": "name: docs\n\non:\n  push:\n    branches: [ main ]\n  pull_request:\n    branches: [ main ]\n  release:\n    types: [ published "
  },
  {
    "path": ".gitignore",
    "chars": 2472,
    "preview": "# Byte-compiled / optimized / DLL files\n__pycache__/\n*.py[cod]\n*$py.class\n\n# C extensions\n*.so\n\n# Distribution / packagi"
  },
  {
    "path": ".pre-commit-config.yaml",
    "chars": 638,
    "preview": "repos:\n-   repo: https://github.com/pre-commit/pre-commit-hooks\n    rev: v4.1.0\n    hooks:\n    -   id: trailing-whitespa"
  },
  {
    "path": "CODEOWNERS",
    "chars": 139,
    "preview": "# Comment line immediately above ownership line is reserved for related gus information. Please be careful while editing"
  },
  {
    "path": "CODE_OF_CONDUCT.md",
    "chars": 5154,
    "preview": "# Salesforce Open Source Community Code of Conduct\n\n## About the Code of Conduct\n\nEquality is a core value at Salesforce"
  },
  {
    "path": "LICENSE.txt",
    "chars": 1502,
    "preview": "BSD 3-Clause License\n\nCopyright (c) 2022 Salesforce, Inc.\nAll rights reserved.\n\nRedistribution and use in source and bin"
  },
  {
    "path": "MANIFEST.in",
    "chars": 267,
    "preview": "recursive-include lavis/configs *.yaml *.json\nrecursive-include lavis/projects *.yaml *.json\n\nrecursive-exclude lavis/da"
  },
  {
    "path": "README.md",
    "chars": 20113,
    "preview": "<p align=\"center\">\n    <br>\n    <img src=\"docs/_static/logo_final.png\" width=\"400\"/>\n    <br>\n<p>\n\n<div align=\"center\">\n"
  },
  {
    "path": "SECURITY.md",
    "chars": 400,
    "preview": "## Security\n\nPlease report any security issue to [security@salesforce.com](mailto:security@salesforce.com)\nas soon as it"
  },
  {
    "path": "app/__init__.py",
    "chars": 666,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "app/calculate_coco_features.py",
    "chars": 2380,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "app/caption.py",
    "chars": 2771,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "app/classification.py",
    "chars": 8076,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "app/dataset_browser.py",
    "chars": 7375,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "app/image_text_match.py",
    "chars": 2825,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "app/main.py",
    "chars": 819,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "app/multimodal_search.py",
    "chars": 7818,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "app/multipage.py",
    "chars": 1318,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "app/text_localization.py",
    "chars": 3457,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "app/utils.py",
    "chars": 2226,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "app/vqa.py",
    "chars": 1967,
    "preview": "\"\"\"\n # Copyright (c) 2022, salesforce.com, inc.\n # All rights reserved.\n # SPDX-License-Identifier: BSD-3-Clause\n # For "
  },
  {
    "path": "dataset_card/avsd_dialogue.md",
    "chars": 1873,
    "preview": "![Samples from the AVSD dataset (Image credit: \"https://arxiv.org/pdf/1901.09107.pdf\").](imgs/avsd_dialogue.png)(Samples"
  },
  {
    "path": "dataset_card/coco_caption.md",
    "chars": 3848,
    "preview": "![Samples from the COCO Caption dataset (Image credit: \"https://arxiv.org/pdf/1504.00325.pdf\").](imgs/coco_caption.png)("
  },
  {
    "path": "dataset_card/coco_retrieval.md",
    "chars": 4037,
    "preview": "![Samples from the COCO Caption dataset (Image credit: \"https://arxiv.org/pdf/1504.00325.pdf\").](imgs/coco_caption.png)("
  },
  {
    "path": "dataset_card/conceptual_captions.md",
    "chars": 2181,
    "preview": "![From https://arxiv.org/pdf/1505.00468.pdf.](imgs/conceptual_captions.png)\n(image credit: https://ai.google.com/researc"
  },
  {
    "path": "dataset_card/didemo_retrieval.md",
    "chars": 4041,
    "preview": "![Samples from the DiDeMo dataset.](imgs/didemo.png)(Samples from the DiDeMo dataset. Image credit: \"https://www.di.ens."
  },
  {
    "path": "dataset_card/flickr_retrieval.md",
    "chars": 3771,
    "preview": "![Samples from Flickr30k dataset (Image credit: \"https://bryanplummer.com/Flickr30kEntities/\").](imgs/flickr30k.png)Samp"
  },
  {
    "path": "dataset_card/gqa.md",
    "chars": 909,
    "preview": "![From https://arxiv.org/abs/1902.09506.pdf.](imgs/gqa.png)\n\n# GQA Dataset\n\n## Description\n(from https://cs.stanford.edu"
  },
  {
    "path": "dataset_card/msrvtt_qa.md",
    "chars": 3439,
    "preview": "![Samples from MSRVTT-QA dataset.](imgs/msrvtt_qa.png)(Samples from MSRVTT-QA dataset, image credit: http://staff.ustc.e"
  },
  {
    "path": "dataset_card/msrvtt_retrieval.md",
    "chars": 2375,
    "preview": "![Samples from Flickr30k dataset (Image credit: \"https://bryanplummer.com/Flickr30kEntities/\").](imgs/msrvtt.png)\n\n# MSR"
  },
  {
    "path": "dataset_card/msvd_qa.md",
    "chars": 2984,
    "preview": "![Samples from MSVD-QA dataset.](imgs/msvd_qa.png)(Samples from MSVD-QA dataset, image credit: http://staff.ustc.edu.cn/"
  },
  {
    "path": "dataset_card/nlvr2.md",
    "chars": 3441,
    "preview": "![From https://arxiv.org/pdf/1505.00468.pdf.](imgs/NLVR2.png)\n\n# Natural Language for Visual Reasoning for Real (NLVR2)\n"
  },
  {
    "path": "dataset_card/nocaps.md",
    "chars": 3178,
    "preview": "![Samples from the COCO Caption dataset (Image credit: \"https://arxiv.org/pdf/1504.00325.pdf\").](imgs/nocaps.png)\n\n# Noc"
  },
  {
    "path": "dataset_card/sbu_caption.md",
    "chars": 770,
    "preview": "![sbu caption](imgs/sbu_caption.png)\n(image credit: http://tamaraberg.com/papers/generation_nips2011.pdf)\n\n# SBU Caption"
  },
  {
    "path": "dataset_card/snli_visual_entailment.md",
    "chars": 3352,
    "preview": "![From https://github.com/necla-ml/SNLI-VE.](imgs/snli_ve.png)\n\n# SNLI-VE: Visual Entailment Dataset\n\n## Description\n(fr"
  },
  {
    "path": "dataset_card/vqav2.md",
    "chars": 3420,
    "preview": "![From https://arxiv.org/pdf/1505.00468.pdf.](imgs/vqav2.png)\n\n# Microsoft COCO Dataset (VQAv2)\n\n## Description\n(from ht"
  },
  {
    "path": "docs/Makefile",
    "chars": 638,
    "preview": "# Minimal makefile for Sphinx documentation\n#\n\n# You can set these variables from the command line, and also\n# from the "
  },
  {
    "path": "docs/benchmark.rst",
    "chars": 14221,
    "preview": "Benchmark\n############\n\nWe provide scripts for evaluating and training models on task datasets. The following benchmark "
  },
  {
    "path": "docs/build_docs.sh",
    "chars": 3319,
    "preview": "#!/bin/bash\nset -euo pipefail\n\n# Change to root directory of repo\nDIRNAME=$(cd \"$( dirname \"${BASH_SOURCE[0]}\" )\" &> /de"
  },
  {
    "path": "docs/conf.py",
    "chars": 1974,
    "preview": "# Configuration file for the Sphinx documentation builder.\n#\n# This file only contains a selection of the most common op"
  },
  {
    "path": "docs/getting_started.rst",
    "chars": 10301,
    "preview": "Dataset Zoo\n##################\nLAVIS inherently supports a wide variety of common language-vision datasets by providing "
  },
  {
    "path": "docs/index.rst",
    "chars": 731,
    "preview": ".. LAVIS documentation master file, created by\n   sphinx-quickstart on Sun Jul 31 10:32:27 2022.\n   You can adapt this f"
  },
  {
    "path": "docs/intro.rst",
    "chars": 7295,
    "preview": "What is LAVIS?\n####################################\n\nLAVIS is a Python deep learning library for LAnguage-and-VISion res"
  },
  {
    "path": "docs/make.bat",
    "chars": 799,
    "preview": "@ECHO OFF\r\n\r\npushd %~dp0\r\n\r\nREM Command file for Sphinx documentation\r\n\r\nif \"%SPHINXBUILD%\" == \"\" (\r\n\tset SPHINXBUILD=sp"
  },
  {
    "path": "docs/requirements.txt",
    "chars": 91,
    "preview": "GitPython\nipykernel\nnbsphinx==0.8.7\npandoc\nsphinx\nsphinx_autodoc_typehints\nsphinx_rtd_theme"
  },
  {
    "path": "docs/tutorial.configs.rst",
    "chars": 5718,
    "preview": ".. _config:\n\nTraining Models on Task Datasets (Commands and Configurations) \n###########################################"
  },
  {
    "path": "docs/tutorial.datasets.rst",
    "chars": 20543,
    "preview": "Adding Datasets\n################################################\n\nThis is a tutorial on adding a new dataset using ``lav"
  },
  {
    "path": "docs/tutorial.evaluation.rst",
    "chars": 1289,
    "preview": "Evaluating Pre-trained Models on Task Datasets\n###############################################\nLAVIS provides pre-traine"
  },
  {
    "path": "docs/tutorial.models.rst",
    "chars": 10799,
    "preview": "Adding Models\n####################################\n\nThis is a tutorial on adding new models using ``lavis.models`` modul"
  },
  {
    "path": "docs/tutorial.processors.rst",
    "chars": 10575,
    "preview": "Adding Processors\n################################################\n\nThis is a tutorial on adding new processors using ``"
  },
  {
    "path": "docs/tutorial.rst",
    "chars": 225,
    "preview": "Tutorials\n==============================\n\n.. toctree::\n   :maxdepth: 1\n\n   tutorial.evaluation\n   tutorial.training-exam"
  },
  {
    "path": "docs/tutorial.tasks.rst",
    "chars": 6989,
    "preview": "Adding Tasks\n####################################\n\nThis is a tutorial on adding new machine learning tasks using ``lavis"
  },
  {
    "path": "docs/tutorial.training-example.rst",
    "chars": 7134,
    "preview": "Example on Finetuning BLIP on COCO-Captioning\n################################################\n\nTo finetune BLIP model o"
  },
  {
    "path": "evaluate.py",
    "chars": 2393,
    "preview": "\"\"\"\n Copyright (c) 2022, salesforce.com, inc.\n All rights reserved.\n SPDX-License-Identifier: BSD-3-Clause\n For full lic"
  },
  {
    "path": "examples/albef_feature_extraction.ipynb",
    "chars": 508093,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": 1,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": [\n "
  },
  {
    "path": "examples/albef_vqa.ipynb",
    "chars": 1631503,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": 1,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": [\n "
  },
  {
    "path": "examples/albef_zero_shot_classification.ipynb",
    "chars": 507969,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": 1,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": [\n "
  },
  {
    "path": "examples/blip2_feature_extraction.ipynb",
    "chars": 3532,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": null,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": "
  },
  {
    "path": "examples/blip2_image_text_matching.ipynb",
    "chars": 3301,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": null,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": "
  },
  {
    "path": "examples/blip2_instructed_generation.ipynb",
    "chars": 511084,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"markdown\",\n   \"metadata\": {},\n   \"source\": [\n    \"#### Large RAM is required to load "
  },
  {
    "path": "examples/blip_feature_extraction.ipynb",
    "chars": 507895,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": 1,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": [\n "
  },
  {
    "path": "examples/blip_image_captioning.ipynb",
    "chars": 507838,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": 1,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": [\n "
  },
  {
    "path": "examples/blip_image_text_matching.ipynb",
    "chars": 507491,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": 17,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": [\n"
  },
  {
    "path": "examples/blip_text_localization.ipynb",
    "chars": 1863151,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": 1,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": [\n "
  },
  {
    "path": "examples/blip_vqa.ipynb",
    "chars": 509204,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": 1,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": [\n "
  },
  {
    "path": "examples/blip_zero_shot_classification.ipynb",
    "chars": 508188,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": 33,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": [\n"
  },
  {
    "path": "examples/clip_feature_extraction.ipynb",
    "chars": 507287,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": 11,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": [\n"
  },
  {
    "path": "examples/clip_zero_shot_classification.ipynb",
    "chars": 508032,
    "preview": "{\n \"cells\": [\n  {\n   \"cell_type\": \"code\",\n   \"execution_count\": 11,\n   \"metadata\": {},\n   \"outputs\": [],\n   \"source\": [\n"
  },
  {
    "path": "lavis/__init__.py",
    "chars": 930,
    "preview": "\"\"\"\n Copyright (c) 2022, salesforce.com, inc.\n All rights reserved.\n SPDX-License-Identifier: BSD-3-Clause\n For full lic"
  },
  {
    "path": "lavis/common/annotator/canny/__init__.py",
    "chars": 155,
    "preview": "import cv2\n\n\nclass CannyDetector:\n    def __call__(self, img, low_threshold, high_threshold):\n        return cv2.Canny(i"
  },
  {
    "path": "lavis/common/annotator/ckpts/download.sh",
    "chars": 222,
    "preview": "#! /bin/bash\n\nwget https://huggingface.co/lllyasviel/ControlNet/resolve/main/annotator/ckpts/dpt_hybrid-midas-501f0c75.p"
  },
  {
    "path": "lavis/common/annotator/hed/__init__.py",
    "chars": 6585,
    "preview": "import numpy as np\nimport cv2\nimport os\nimport torch\nfrom einops import rearrange\nfrom annotator.util import annotator_c"
  },
  {
    "path": "lavis/common/annotator/midas/__init__.py",
    "chars": 1400,
    "preview": "import cv2\nimport numpy as np\nimport torch\n\nfrom einops import rearrange\nfrom .api import MiDaSInference\n\n\nclass MidasDe"
  },
  {
    "path": "lavis/common/annotator/midas/api.py",
    "chars": 5229,
    "preview": "# based on https://github.com/isl-org/MiDaS\n\nimport cv2\nimport os\nimport torch\nimport torch.nn as nn\nfrom torchvision.tr"
  },
  {
    "path": "lavis/common/annotator/midas/midas/__init__.py",
    "chars": 0,
    "preview": ""
  },
  {
    "path": "lavis/common/annotator/midas/midas/base_model.py",
    "chars": 367,
    "preview": "import torch\n\n\nclass BaseModel(torch.nn.Module):\n    def load(self, path):\n        \"\"\"Load model from file.\n\n        Arg"
  },
  {
    "path": "lavis/common/annotator/midas/midas/blocks.py",
    "chars": 9242,
    "preview": "import torch\nimport torch.nn as nn\n\nfrom .vit import (\n    _make_pretrained_vitb_rn50_384,\n    _make_pretrained_vitl16_3"
  },
  {
    "path": "lavis/common/annotator/midas/midas/dpt_depth.py",
    "chars": 3154,
    "preview": "import torch\nimport torch.nn as nn\nimport torch.nn.functional as F\n\nfrom .base_model import BaseModel\nfrom .blocks impor"
  },
  {
    "path": "lavis/common/annotator/midas/midas/midas_net.py",
    "chars": 2709,
    "preview": "\"\"\"MidashNet: Network for monocular depth estimation trained by mixing several datasets.\nThis file contains code that is"
  },
  {
    "path": "lavis/common/annotator/midas/midas/midas_net_custom.py",
    "chars": 5207,
    "preview": "\"\"\"MidashNet: Network for monocular depth estimation trained by mixing several datasets.\nThis file contains code that is"
  },
  {
    "path": "lavis/common/annotator/midas/midas/transforms.py",
    "chars": 7869,
    "preview": "import numpy as np\nimport cv2\nimport math\n\n\ndef apply_min_size(sample, size, image_interpolation_method=cv2.INTER_AREA):"
  },
  {
    "path": "lavis/common/annotator/midas/midas/vit.py",
    "chars": 14625,
    "preview": "import torch\nimport torch.nn as nn\nimport timm\nimport types\nimport math\nimport torch.nn.functional as F\n\n\nclass Slice(nn"
  },
  {
    "path": "lavis/common/annotator/midas/utils.py",
    "chars": 4582,
    "preview": "\"\"\"Utils for monoDepth.\"\"\"\nimport sys\nimport re\nimport numpy as np\nimport cv2\nimport torch\n\n\ndef read_pfm(path):\n    \"\"\""
  },
  {
    "path": "lavis/common/annotator/mlsd/__init__.py",
    "chars": 1458,
    "preview": "import cv2\nimport numpy as np\nimport torch\nimport os\n\nfrom einops import rearrange\nfrom .models.mbv2_mlsd_tiny import Mo"
  },
  {
    "path": "lavis/common/annotator/mlsd/models/mbv2_mlsd_large.py",
    "chars": 9678,
    "preview": "import os\nimport sys\nimport torch\nimport torch.nn as nn\nimport torch.utils.model_zoo as model_zoo\nfrom  torch.nn import "
  },
  {
    "path": "lavis/common/annotator/mlsd/models/mbv2_mlsd_tiny.py",
    "chars": 9180,
    "preview": "import os\nimport sys\nimport torch\nimport torch.nn as nn\nimport torch.utils.model_zoo as model_zoo\nfrom  torch.nn import "
  },
  {
    "path": "lavis/common/annotator/mlsd/utils.py",
    "chars": 24049,
    "preview": "'''\nmodified by  lihaoweicv\npytorch version\n'''\n\n'''\nM-LSD\nCopyright 2021-present NAVER Corp.\nApache License v2.0\n'''\n\ni"
  },
  {
    "path": "lavis/common/annotator/openpose/__init__.py",
    "chars": 1957,
    "preview": "import os\nos.environ[\"KMP_DUPLICATE_LIB_OK\"]=\"TRUE\"\n\nimport torch\nimport numpy as np\nfrom . import util\nfrom .body impor"
  },
  {
    "path": "lavis/common/annotator/openpose/body.py",
    "chars": 10994,
    "preview": "import cv2\nimport numpy as np\nimport math\nimport time\nfrom scipy.ndimage.filters import gaussian_filter\nimport matplotli"
  },
  {
    "path": "lavis/common/annotator/openpose/hand.py",
    "chars": 3426,
    "preview": "import cv2\nimport json\nimport numpy as np\nimport math\nimport time\nfrom scipy.ndimage.filters import gaussian_filter\nimpo"
  },
  {
    "path": "lavis/common/annotator/openpose/model.py",
    "chars": 8745,
    "preview": "import torch\nfrom collections import OrderedDict\n\nimport torch\nimport torch.nn as nn\n\ndef make_layers(block, no_relu_lay"
  },
  {
    "path": "lavis/common/annotator/openpose/util.py",
    "chars": 7507,
    "preview": "import math\nimport numpy as np\nimport matplotlib\nimport cv2\n\n\ndef padRightDownCorner(img, stride, padValue):\n    h = img"
  },
  {
    "path": "lavis/common/annotator/uniformer/__init__.py",
    "chars": 1070,
    "preview": "import os\n\nfrom annotator.uniformer.mmseg.apis import init_segmentor, inference_segmentor, show_result_pyplot\nfrom annot"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/datasets/ade20k.py",
    "chars": 1844,
    "preview": "# dataset settings\ndataset_type = 'ADE20KDataset'\ndata_root = 'data/ade/ADEChallengeData2016'\nimg_norm_cfg = dict(\n    m"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/datasets/chase_db1.py",
    "chars": 1924,
    "preview": "# dataset settings\ndataset_type = 'ChaseDB1Dataset'\ndata_root = 'data/CHASE_DB1'\nimg_norm_cfg = dict(\n    mean=[123.675,"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/datasets/cityscapes.py",
    "chars": 1780,
    "preview": "# dataset settings\ndataset_type = 'CityscapesDataset'\ndata_root = 'data/cityscapes/'\nimg_norm_cfg = dict(\n    mean=[123."
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/datasets/cityscapes_769x769.py",
    "chars": 1281,
    "preview": "_base_ = './cityscapes.py'\nimg_norm_cfg = dict(\n    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb="
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/datasets/drive.py",
    "chars": 1915,
    "preview": "# dataset settings\ndataset_type = 'DRIVEDataset'\ndata_root = 'data/DRIVE'\nimg_norm_cfg = dict(\n    mean=[123.675, 116.28"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/datasets/hrf.py",
    "chars": 1915,
    "preview": "# dataset settings\ndataset_type = 'HRFDataset'\ndata_root = 'data/HRF'\nimg_norm_cfg = dict(\n    mean=[123.675, 116.28, 10"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/datasets/pascal_context.py",
    "chars": 1998,
    "preview": "# dataset settings\ndataset_type = 'PascalContextDataset'\ndata_root = 'data/VOCdevkit/VOC2010/'\nimg_norm_cfg = dict(\n    "
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/datasets/pascal_context_59.py",
    "chars": 2024,
    "preview": "# dataset settings\ndataset_type = 'PascalContextDataset59'\ndata_root = 'data/VOCdevkit/VOC2010/'\nimg_norm_cfg = dict(\n  "
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/datasets/pascal_voc12.py",
    "chars": 1930,
    "preview": "# dataset settings\ndataset_type = 'PascalVOCDataset'\ndata_root = 'data/VOCdevkit/VOC2012'\nimg_norm_cfg = dict(\n    mean="
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/datasets/pascal_voc12_aug.py",
    "chars": 261,
    "preview": "_base_ = './pascal_voc12.py'\n# dataset settings\ndata = dict(\n    train=dict(\n        ann_dir=['SegmentationClass', 'Segm"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/datasets/stare.py",
    "chars": 1917,
    "preview": "# dataset settings\ndataset_type = 'STAREDataset'\ndata_root = 'data/STARE'\nimg_norm_cfg = dict(\n    mean=[123.675, 116.28"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/default_runtime.py",
    "chars": 321,
    "preview": "# yapf:disable\nlog_config = dict(\n    interval=50,\n    hooks=[\n        dict(type='TextLoggerHook', by_epoch=False),\n    "
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/ann_r50-d8.py",
    "chars": 1346,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/apcnet_r50-d8.py",
    "chars": 1302,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/ccnet_r50-d8.py",
    "chars": 1258,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/cgnet.py",
    "chars": 1110,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', eps=1e-03, requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/danet_r50-d8.py",
    "chars": 1261,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/deeplabv3_r50-d8.py",
    "chars": 1273,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/deeplabv3_unet_s5-d16.py",
    "chars": 1499,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/deeplabv3plus_r50-d8.py",
    "chars": 1343,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/dmnet_r50-d8.py",
    "chars": 1302,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/dnl_r50-d8.py",
    "chars": 1316,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/emanet_r50-d8.py",
    "chars": 1329,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/encnet_r50-d8.py",
    "chars": 1435,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/fast_scnn.py",
    "chars": 1761,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True, momentum=0.01)\nmodel = dict(\n    type='EncoderDecode"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/fcn_hr18.py",
    "chars": 1646,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/fcn_r50-d8.py",
    "chars": 1285,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/fcn_unet_s5-d16.py",
    "chars": 1512,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/fpn_r50.py",
    "chars": 1056,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/fpn_uniformer.py",
    "chars": 977,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    backbon"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/gcnet_r50-d8.py",
    "chars": 1326,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/lraspp_m-v3-d8.py",
    "chars": 766,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', eps=0.001, requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/nonlocal_r50-d8.py",
    "chars": 1315,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/ocrnet_hr18.py",
    "chars": 2196,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='CascadeEncoderDecoder',\n    "
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/ocrnet_r50-d8.py",
    "chars": 1385,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='CascadeEncoderDecoder',\n    "
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/pointrend_r50.py",
    "chars": 1704,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='CascadeEncoderDecoder',\n    "
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/psanet_r50-d8.py",
    "chars": 1406,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/pspnet_r50-d8.py",
    "chars": 1271,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/pspnet_unet_s5-d16.py",
    "chars": 1497,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/upernet_r50.py",
    "chars": 1301,
    "preview": "# model settings\nnorm_cfg = dict(type='SyncBN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrai"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/models/upernet_uniformer.py",
    "chars": 1235,
    "preview": "# model settings\nnorm_cfg = dict(type='BN', requires_grad=True)\nmodel = dict(\n    type='EncoderDecoder',\n    pretrained="
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/schedules/schedule_160k.py",
    "chars": 382,
    "preview": "# optimizer\noptimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005)\noptimizer_config = dict()\n# learnin"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/schedules/schedule_20k.py",
    "chars": 379,
    "preview": "# optimizer\noptimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005)\noptimizer_config = dict()\n# learnin"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/schedules/schedule_40k.py",
    "chars": 379,
    "preview": "# optimizer\noptimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005)\noptimizer_config = dict()\n# learnin"
  },
  {
    "path": "lavis/common/annotator/uniformer/configs/_base_/schedules/schedule_80k.py",
    "chars": 379,
    "preview": "# optimizer\noptimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005)\noptimizer_config = dict()\n# learnin"
  },
  {
    "path": "lavis/common/annotator/uniformer/exp/upernet_global_small/config.py",
    "chars": 1316,
    "preview": "_base_ = [\n    '../../configs/_base_/models/upernet_uniformer.py', \n    '../../configs/_base_/datasets/ade20k.py',\n    '"
  },
  {
    "path": "lavis/common/annotator/uniformer/exp/upernet_global_small/run.sh",
    "chars": 382,
    "preview": "#!/usr/bin/env bash\n\nwork_path=$(dirname $0)\nPYTHONPATH=\"$(dirname $0)/../../\":$PYTHONPATH \\\npython -m torch.distributed"
  },
  {
    "path": "lavis/common/annotator/uniformer/exp/upernet_global_small/test.sh",
    "chars": 318,
    "preview": "#!/usr/bin/env bash\n\nwork_path=$(dirname $0)\nPYTHONPATH=\"$(dirname $0)/../../\":$PYTHONPATH \\\npython -m torch.distributed"
  },
  {
    "path": "lavis/common/annotator/uniformer/exp/upernet_global_small/test_config_g.py",
    "chars": 1317,
    "preview": "_base_ = [\n    '../../configs/_base_/models/upernet_uniformer.py', \n    '../../configs/_base_/datasets/ade20k.py',\n    '"
  },
  {
    "path": "lavis/common/annotator/uniformer/exp/upernet_global_small/test_config_h32.py",
    "chars": 1339,
    "preview": "_base_ = [\n    '../../configs/_base_/models/upernet_uniformer.py', \n    '../../configs/_base_/datasets/ade20k.py',\n    '"
  },
  {
    "path": "lavis/common/annotator/uniformer/exp/upernet_global_small/test_config_w32.py",
    "chars": 1339,
    "preview": "_base_ = [\n    '../../configs/_base_/models/upernet_uniformer.py', \n    '../../configs/_base_/datasets/ade20k.py',\n    '"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/__init__.py",
    "chars": 352,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\n# flake8: noqa\nfrom .arraymisc import *\nfrom .fileio import *\nfrom .imag"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/arraymisc/__init__.py",
    "chars": 133,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom .quantization import dequantize, quantize\n\n__all__ = ['quantize', '"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/arraymisc/quantization.py",
    "chars": 1824,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport numpy as np\n\n\ndef quantize(arr, min_val, max_val, levels, dtype=n"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/__init__.py",
    "chars": 2438,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom .alexnet import AlexNet\n# yapf: disable\nfrom .bricks import (ACTIVA"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/alexnet.py",
    "chars": 1990,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport logging\n\nimport torch.nn as nn\n\n\nclass AlexNet(nn.Module):\n    \"\""
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/__init__.py",
    "chars": 1732,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom .activation import build_activation_layer\nfrom .context_block impor"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/activation.py",
    "chars": 2508,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\n\nfrom"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/context_block.py",
    "chars": 4681,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch\nfrom torch import nn\n\nfrom ..utils import constant_init, ka"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/conv.py",
    "chars": 1446,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom torch import nn\n\nfrom .registry import CONV_LAYERS\n\nCONV_LAYERS.reg"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/conv2d_adaptive_padding.py",
    "chars": 2514,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport math\n\nfrom torch import nn\nfrom torch.nn import functional as F\n\n"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/conv_module.py",
    "chars": 8760,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport warnings\n\nimport torch.nn as nn\n\nfrom annotator.uniformer.mmcv.ut"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/conv_ws.py",
    "chars": 5417,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\n\nfrom"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/depthwise_separable_conv_module.py",
    "chars": 4142,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch.nn as nn\n\nfrom .conv_module import ConvModule\n\n\nclass Depth"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/drop.py",
    "chars": 2172,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch\nimport torch.nn as nn\n\nfrom annotator.uniformer.mmcv import"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/generalized_attention.py",
    "chars": 15999,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport math\n\nimport numpy as np\nimport torch\nimport torch.nn as nn\nimpor"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/hsigmoid.py",
    "chars": 1097,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch.nn as nn\n\nfrom .registry import ACTIVATION_LAYERS\n\n\n@ACTIVA"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/hswish.py",
    "chars": 651,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch.nn as nn\n\nfrom .registry import ACTIVATION_LAYERS\n\n\n@ACTIVA"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/non_local.py",
    "chars": 11012,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom abc import ABCMeta\n\nimport torch\nimport torch.nn as nn\n\nfrom ..util"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/norm.py",
    "chars": 5154,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport inspect\n\nimport torch.nn as nn\n\nfrom annotator.uniformer.mmcv.uti"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/padding.py",
    "chars": 1127,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch.nn as nn\n\nfrom .registry import PADDING_LAYERS\n\nPADDING_LAY"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/plugin.py",
    "chars": 2487,
    "preview": "import inspect\nimport platform\n\nfrom .registry import PLUGIN_LAYERS\n\nif platform.system() == 'Windows':\n    import regex"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/registry.py",
    "chars": 658,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom annotator.uniformer.mmcv.utils import Registry\n\nCONV_LAYERS = Regis"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/scale.py",
    "chars": 577,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch\nimport torch.nn as nn\n\n\nclass Scale(nn.Module):\n    \"\"\"A le"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/swish.py",
    "chars": 485,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch\nimport torch.nn as nn\n\nfrom .registry import ACTIVATION_LAY"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/transformer.py",
    "chars": 24637,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport copy\nimport warnings\n\nimport torch\nimport torch.nn as nn\n\nfrom an"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/upsample.py",
    "chars": 2880,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch.nn as nn\nimport torch.nn.functional as F\n\nfrom ..utils impo"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/bricks/wrappers.py",
    "chars": 6961,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nr\"\"\"Modified from https://github.com/facebookresearch/detectron2/blob/ma"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/builder.py",
    "chars": 1089,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom ..runner import Sequential\nfrom ..utils import Registry, build_from"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/resnet.py",
    "chars": 9955,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport logging\n\nimport torch.nn as nn\nimport torch.utils.checkpoint as c"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/utils/__init__.py",
    "chars": 1023,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom .flops_counter import get_model_complexity_info\nfrom .fuse_conv_bn "
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/utils/flops_counter.py",
    "chars": 22104,
    "preview": "# Modified from flops-counter.pytorch by Vladislav Sovrasov\n# original repo: https://github.com/sovrasov/flops-counter.p"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/utils/fuse_conv_bn.py",
    "chars": 1881,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport torch\nimport torch.nn as nn\n\n\ndef _fuse_conv_bn(conv, bn):\n    \"\""
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/utils/sync_bn.py",
    "chars": 2327,
    "preview": "import torch\n\nimport annotator.uniformer.mmcv as mmcv\n\n\nclass _BatchNormXd(torch.nn.modules.batchnorm._BatchNorm):\n    \""
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/utils/weight_init.py",
    "chars": 26006,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport copy\nimport math\nimport warnings\n\nimport numpy as np\nimport torch"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/cnn/vgg.py",
    "chars": 6053,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport logging\n\nimport torch.nn as nn\n\nfrom .utils import constant_init,"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/engine/__init__.py",
    "chars": 266,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom .test import (collect_results_cpu, collect_results_gpu, multi_gpu_t"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/engine/test.py",
    "chars": 7196,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport os.path as osp\nimport pickle\nimport shutil\nimport tempfile\nimport"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/fileio/__init__.py",
    "chars": 478,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom .file_client import BaseStorageBackend, FileClient\nfrom .handlers i"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/fileio/file_client.py",
    "chars": 41933,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport inspect\nimport os\nimport os.path as osp\nimport re\nimport tempfile"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/fileio/handlers/__init__.py",
    "chars": 278,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom .base import BaseFileHandler\nfrom .json_handler import JsonHandler\n"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/fileio/handlers/base.py",
    "chars": 993,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom abc import ABCMeta, abstractmethod\n\n\nclass BaseFileHandler(metaclas"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/fileio/handlers/json_handler.py",
    "chars": 1068,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport json\n\nimport numpy as np\n\nfrom .base import BaseFileHandler\n\n\ndef"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/fileio/handlers/pickle_handler.py",
    "chars": 817,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport pickle\n\nfrom .base import BaseFileHandler\n\n\nclass PickleHandler(B"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/fileio/handlers/yaml_handler.py",
    "chars": 665,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport yaml\n\ntry:\n    from yaml import CLoader as Loader, CDumper as Dum"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/fileio/io.py",
    "chars": 5520,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom io import BytesIO, StringIO\nfrom pathlib import Path\n\nfrom ..utils "
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/fileio/parse.py",
    "chars": 3458,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\n\nfrom io import StringIO\n\nfrom .file_client import FileClient\n\n\ndef list"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/image/__init__.py",
    "chars": 1725,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom .colorspace import (bgr2gray, bgr2hls, bgr2hsv, bgr2rgb, bgr2ycbcr,"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/image/colorspace.py",
    "chars": 9907,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport cv2\nimport numpy as np\n\n\ndef imconvert(img, src, dst):\n    \"\"\"Con"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/image/geometric.py",
    "chars": 25196,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport numbers\n\nimport cv2\nimport numpy as np\n\nfrom ..utils import to_2t"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/image/io.py",
    "chars": 9572,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport io\nimport os.path as osp\nfrom pathlib import Path\n\nimport cv2\nimp"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/image/misc.py",
    "chars": 1410,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport numpy as np\n\nimport annotator.uniformer.mmcv as mmcv\n\ntry:\n    im"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/image/photometric.py",
    "chars": 14999,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nimport cv2\nimport numpy as np\n\nfrom ..utils import is_tuple_of\nfrom .col"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/model_zoo/deprecated.json",
    "chars": 217,
    "preview": "{\n  \"resnet50_caffe\": \"detectron/resnet50_caffe\",\n  \"resnet50_caffe_bgr\": \"detectron2/resnet50_caffe_bgr\",\n  \"resnet101_"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/model_zoo/mmcls.json",
    "chars": 3690,
    "preview": "{\n  \"vgg11\": \"https://download.openmmlab.com/mmclassification/v0/vgg/vgg11_batch256_imagenet_20210208-4271cd6c.pth\",\n  \""
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/model_zoo/open_mmlab.json",
    "chars": 5181,
    "preview": "{\n  \"vgg16_caffe\": \"https://download.openmmlab.com/pretrain/third_party/vgg16_caffe-292e1171.pth\",\n  \"detectron/resnet50"
  },
  {
    "path": "lavis/common/annotator/uniformer/mmcv/ops/__init__.py",
    "chars": 4506,
    "preview": "# Copyright (c) OpenMMLab. All rights reserved.\nfrom .assign_score_withk import assign_score_withk\nfrom .ball_query impo"
  }
]

// ... and 1183 more files (download for full content)