repo_full_name
stringlengths
6
93
repo_url
stringlengths
25
112
repo_api_url
stringclasses
28 values
owner
stringclasses
28 values
repo_name
stringclasses
28 values
description
stringclasses
28 values
stars
int64
617
98.8k
forks
int64
31
355
watchers
int64
990
999
license
stringclasses
2 values
default_branch
stringclasses
2 values
repo_created_at
timestamp[s]date
2012-07-24 23:12:50
2025-06-16 08:07:28
repo_updated_at
timestamp[s]date
2026-02-23 15:23:15
2026-05-03 18:52:12
repo_topics
listlengths
0
13
repo_languages
unknown
is_fork
bool
1 class
open_issues
int64
3
104
file_path
stringlengths
3
208
file_name
stringclasses
509 values
file_extension
stringclasses
1 value
file_size_bytes
int64
101
84k
file_url
stringclasses
627 values
file_raw_url
stringclasses
627 values
file_sha
stringclasses
624 values
language
stringclasses
8 values
parsed_at
stringdate
2026-05-04 01:12:36
2026-05-04 19:41:55
text
stringlengths
100
102k
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/datasets.py
null
null
null
null
null
null
Python
2026-05-04T02:49:54.323755
# Copyright (c) 2015-present, Facebook, Inc. # All rights reserved. import os import json from torchvision import datasets, transforms from torchvision.datasets.folder import ImageFolder, default_loader from timm.data.constants import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD from timm.data import create_transform ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/engine.py
null
null
null
null
null
null
Python
2026-05-04T02:49:54.345390
# Copyright (c) 2015-present, Facebook, Inc. # All rights reserved. """ Train and eval functions used in main.py """ import math import sys from typing import Iterable, Optional import torch from timm.data import Mixup from timm.utils import accuracy, ModelEma from losses import DistillationLoss import utils def t...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/losses.py
null
null
null
null
null
null
Python
2026-05-04T02:49:54.705670
# Copyright (c) 2015-present, Facebook, Inc. # All rights reserved. """ Implements the knowledge distillation loss """ import torch from torch.nn import functional as F class DistillationLoss(torch.nn.Module): """ This module wraps a standard criterion and adds an extra knowledge distillation loss by taki...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/main.py
null
null
null
null
null
null
Python
2026-05-04T02:49:54.707104
# Copyright (c) 2015-present, Facebook, Inc. # All rights reserved. import argparse import datetime import numpy as np import time import torch import torch.backends.cudnn as cudnn import json from pathlib import Path from timm.data import Mixup from timm.models import create_model from timm.loss import LabelSmoothin...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/get_flops.py
null
null
null
null
null
null
Python
2026-05-04T02:49:54.708652
import argparse import torch from timm.models import create_model import pvt import pvt_v2 try: from mmcv.cnn import get_model_complexity_info from mmcv.cnn.utils.flops_counter import get_model_complexity_info, flops_to_string, params_to_string except ImportError: raise ImportError('Please upgrade mmcv to ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/mcloader/data_prefetcher.py
null
null
null
null
null
null
Python
2026-05-04T02:49:54.710186
import torch class DataPrefetcher: def __init__(self, loader): self.loader = iter(loader) self.stream = torch.cuda.Stream() self.preload() def preload(self): try: self.next_input, self.next_target = next(self.loader) except StopIteration: self.n...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/mcloader/classification.py
null
null
null
null
null
null
Python
2026-05-04T02:49:54.711167
import torch from torch.utils.data import Dataset from .imagenet import ImageNet class ClassificationDataset(Dataset): """Dataset for classification. """ def __init__(self, split='train', pipeline=None): if split == 'train': self.data_source = ImageNet(root='data/imagenet/train', ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/hubconf.py
null
null
null
null
null
null
Python
2026-05-04T02:49:55.134308
# Copyright (c) 2015-present, Facebook, Inc. # All rights reserved. from models import * dependencies = ["torch", "torchvision", "timm"]
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/mcloader/imagenet.py
null
null
null
null
null
null
Python
2026-05-04T02:49:55.444987
from .image_list import ImageList class ImageNet(ImageList): def __init__(self, root, list_file, memcached, mclient_path): super(ImageNet, self).__init__( root, list_file, memcached, mclient_path)
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/mcloader/mcloader.py
null
null
null
null
null
null
Python
2026-05-04T02:49:55.453396
import io from PIL import Image try: import mc except ImportError as E: pass def pil_loader(img_str): buff = io.BytesIO(img_str) return Image.open(buff) class McLoader(object): def __init__(self, mclient_path): assert mclient_path is not None, \ "Please specify 'data_mclient...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/mcloader/image_list.py
null
null
null
null
null
null
Python
2026-05-04T02:49:55.457752
import os from PIL import Image from .mcloader import McLoader class ImageList(object): def __init__(self, root, list_file, memcached=False, mclient_path=None): with open(list_file, 'r') as f: lines = f.readlines() self.has_labels = len(lines[0].split()) == 2 if self.has_labe...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/pvt_v2.py
null
null
null
null
null
null
Python
2026-05-04T02:49:55.575723
import torch import torch.nn as nn import torch.nn.functional as F from functools import partial from timm.models.layers import DropPath, to_2tuple, trunc_normal_ from timm.models.registry import register_model from timm.models.vision_transformer import _cfg import math class Mlp(nn.Module): def __init__(self, i...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/samplers.py
null
null
null
null
null
null
Python
2026-05-04T02:49:55.587834
# Copyright (c) 2015-present, Facebook, Inc. # All rights reserved. import torch import torch.distributed as dist import math class RASampler(torch.utils.data.Sampler): """Sampler that restricts data loading to a subset of the dataset for distributed, with repeated augmentation. It ensures that different ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/analyze_results.py
null
null
null
null
null
null
Python
2026-05-04T02:49:55.589455
import argparse import os.path as osp import mmcv import numpy as np from mmcv import Config, DictAction from mmdet.core.evaluation import eval_map from mmdet.core.visualization import imshow_gt_det_bboxes from mmdet.datasets import build_dataset, get_loading_pipeline def bbox_map_eval(det_result, annotation): ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/utils.py
null
null
null
null
null
null
Python
2026-05-04T02:49:55.593423
# Copyright (c) 2015-present, Facebook, Inc. # All rights reserved. """ Misc functions, including distributed helpers. Mostly copy-paste from torchvision references. """ import io import os import time from collections import defaultdict, deque import datetime import torch import torch.distributed as dist import mmcv...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/run_with_submitit.py
null
null
null
null
null
null
Python
2026-05-04T02:49:55.608526
# Copyright (c) 2015-present, Facebook, Inc. # All rights reserved. """ A script to run multinode training with submitit. """ import argparse import os import os.path as osp import uuid from pathlib import Path import main as classification import submitit def parse_args(): classification_parser = classification...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
classification/pvt.py
null
null
null
null
null
null
Python
2026-05-04T02:49:55.609534
import torch import torch.nn as nn import torch.nn.functional as F from functools import partial from timm.models.layers import DropPath, to_2tuple, trunc_normal_ from timm.models.registry import register_model from timm.models.vision_transformer import _cfg __all__ = [ 'pvt_tiny', 'pvt_small', 'pvt_medium', 'pvt...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/benchmark.py
null
null
null
null
null
null
Python
2026-05-04T02:49:55.722454
import argparse import time import torch from mmcv import Config, DictAction from mmcv.cnn import fuse_conv_bn from mmcv.parallel import MMDataParallel from mmcv.runner import load_checkpoint, wrap_fp16_model from mmdet.datasets import (build_dataloader, build_dataset, replace_ImageToTenso...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/datasets/cityscapes_instance.py
null
null
null
null
null
null
Python
2026-05-04T02:49:56.125633
# dataset settings dataset_type = 'CityscapesDataset' data_root = 'data/cityscapes/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True, with_mask=True), dict( ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/datasets/coco_instance_semantic.py
null
null
null
null
null
null
Python
2026-05-04T02:49:56.253551
# dataset settings dataset_type = 'CocoDataset' data_root = 'data/coco/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict( type='LoadAnnotations', with_bbox=True, with_mask=True, with_seg=True), ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/datasets/cityscapes_detection.py
null
null
null
null
null
null
Python
2026-05-04T02:49:56.254597
# dataset settings dataset_type = 'CityscapesDataset' data_root = 'data/cityscapes/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict( type='Resize'...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/datasets/coco_instance.py
null
null
null
null
null
null
Python
2026-05-04T02:49:56.257608
# dataset settings dataset_type = 'CocoDataset' data_root = 'data/coco/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True, with_mask=True), dict(type='Resize', img...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/datasets/lvis_v1_instance.py
null
null
null
null
null
null
Python
2026-05-04T02:49:56.258528
# dataset settings _base_ = 'coco_instance.py' dataset_type = 'LVISV1Dataset' data_root = 'data/lvis_v1/' data = dict( samples_per_gpu=2, workers_per_gpu=2, train=dict( _delete_=True, type='ClassBalancedDataset', oversample_thr=1e-3, dataset=dict( type=dataset_typ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/datasets/coco_detection.py
null
null
null
null
null
null
Python
2026-05-04T02:49:56.259532
# dataset settings dataset_type = 'CocoDataset' data_root = 'data/coco/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1333, 80...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/datasets/voc0712.py
null
null
null
null
null
null
Python
2026-05-04T02:49:56.260762
# dataset settings dataset_type = 'VOCDataset' data_root = 'data/VOCdevkit/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1000...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/datasets/deepfashion.py
null
null
null
null
null
null
Python
2026-05-04T02:49:56.261736
# dataset settings dataset_type = 'DeepFashionDataset' data_root = 'data/DeepFashion/In-shop/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True, with_mask=True), d...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/datasets/lvis_v0.5_instance.py
null
null
null
null
null
null
Python
2026-05-04T02:49:56.262725
# dataset settings _base_ = 'coco_instance.py' dataset_type = 'LVISV05Dataset' data_root = 'data/lvis_v0.5/' data = dict( samples_per_gpu=2, workers_per_gpu=2, train=dict( _delete_=True, type='ClassBalancedDataset', oversample_thr=1e-3, dataset=dict( type=dataset_...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/datasets/wider_face.py
null
null
null
null
null
null
Python
2026-05-04T02:49:56.545842
# dataset settings dataset_type = 'WIDERFaceDataset' data_root = 'data/WIDERFace/' img_norm_cfg = dict(mean=[123.675, 116.28, 103.53], std=[1, 1, 1], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile', to_float32=True), dict(type='LoadAnnotations', with_bbox=True), dict( type='PhotoMetric...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/models/faster_rcnn_r50_caffe_c4.py
null
null
null
null
null
null
Python
2026-05-04T02:49:57.278342
# model settings norm_cfg = dict(type='BN', requires_grad=False) model = dict( type='FasterRCNN', pretrained='open-mmlab://detectron2/resnet50_caffe', backbone=dict( type='ResNet', depth=50, num_stages=3, strides=(1, 2, 2), dilations=(1, 1, 1), out_indices=(2,...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/models/fast_rcnn_r50_fpn.py
null
null
null
null
null
null
Python
2026-05-04T02:49:57.280121
# model settings model = dict( type='FastRCNN', pretrained='torchvision://resnet50', backbone=dict( type='ResNet', depth=50, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/models/cascade_mask_rcnn_pvtv2_b2_fpn.py
null
null
null
null
null
null
Python
2026-05-04T02:49:57.280626
# model settings model = dict( type='CascadeRCNN', backbone=dict( type='pvt_v2_b2', style='pytorch'), neck=dict( type='FPN', in_channels=[64, 128, 320, 512], out_channels=256, num_outs=5), rpn_head=dict( type='RPNHead', in_channels=256, ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/models/cascade_rcnn_r50_fpn.py
null
null
null
null
null
null
Python
2026-05-04T02:49:57.281871
# model settings model = dict( type='CascadeRCNN', pretrained='torchvision://resnet50', backbone=dict( type='ResNet', depth=50, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/models/faster_rcnn_r50_fpn.py
null
null
null
null
null
null
Python
2026-05-04T02:49:57.282356
# model settings model = dict( type='FasterRCNN', pretrained='torchvision://resnet50', backbone=dict( type='ResNet', depth=50, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/models/cascade_mask_rcnn_r50_fpn.py
null
null
null
null
null
null
Python
2026-05-04T02:49:57.282811
# model settings model = dict( type='CascadeRCNN', pretrained='torchvision://resnet50', backbone=dict( type='ResNet', depth=50, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, ...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/models/faster_rcnn_r50_caffe_dc5.py
null
null
null
null
null
null
Python
2026-05-04T02:49:57.283453
# model settings norm_cfg = dict(type='BN', requires_grad=False) model = dict( type='FasterRCNN', pretrained='open-mmlab://detectron2/resnet50_caffe', backbone=dict( type='ResNet', depth=50, num_stages=4, strides=(1, 2, 2, 1), dilations=(1, 1, 1, 2), out_indic...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/default_runtime.py
null
null
null
null
null
null
Python
2026-05-04T02:49:57.283911
checkpoint_config = dict(interval=1) # yapf:disable log_config = dict( interval=50, hooks=[ dict(type='TextLoggerHook'), # dict(type='TensorboardLoggerHook') ]) # yapf:enable custom_hooks = [dict(type='NumClassCheckHook')] dist_params = dict(backend='nccl') log_level = 'INFO' load_from = No...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/models/mask_rcnn_r50_caffe_c4.py
null
null
null
null
null
null
Python
2026-05-04T02:49:57.306647
# model settings norm_cfg = dict(type='BN', requires_grad=False) model = dict( type='MaskRCNN', pretrained='open-mmlab://detectron2/resnet50_caffe', backbone=dict( type='ResNet', depth=50, num_stages=3, strides=(1, 2, 2), dilations=(1, 1, 1), out_indices=(2, )...
whai362/PVT
https://github.com/whai362/PVT
null
null
null
null
1,893
null
null
apache-2.0
null
null
null
null
null
null
null
detection/configs/_base_/models/mask_rcnn_r50_fpn.py
null
null
null
null
null
null
Python
2026-05-04T02:49:57.511049
# model settings model = dict( type='MaskRCNN', pretrained='torchvision://resnet50', backbone=dict( type='ResNet', depth=50, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, ...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
benchmark/dataset_extract/dataset_extract_kitti.py
null
null
null
null
null
null
Python
2026-05-04T02:50:00.281811
import os import numpy as np import os.path as osp from PIL import Image from tqdm import tqdm import csv import cv2 import json import glob import shutil from natsort import natsorted from eval_utils import even_or_odd from eval_utils import gen_json, get_sorted_files, copy_crop_files def extract_kitti( root, ...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
benchmark/dataset_extract/dataset_extract_sintel.py
null
null
null
null
null
null
Python
2026-05-04T02:50:00.296320
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. # # Data loading based on https://github.com/NVIDIA/flownet2-pytorch import os import numpy as np import os.path as osp f...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
benchmark/dataset_extract/dataset_extract_bonn.py
null
null
null
null
null
null
Python
2026-05-04T02:50:00.298463
import os import numpy as np import os.path as osp from PIL import Image from tqdm import tqdm import cv2 import csv import json import glob import shutil from natsort import natsorted from eval_utils import gen_json, get_sorted_files, even_or_odd, copy_crop_files def extract_bonn( root, depth_root, saved...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
benchmark/dataset_extract/eval_utils.py
null
null
null
null
null
null
Python
2026-05-04T02:50:00.299889
import os import numpy as np import os.path as osp import json import glob import cv2 import shutil from PIL import Image from natsort import natsorted def even_or_odd(num): if num % 2 == 0: return num else: return num - 1 def gen_json(root_path, dataset, start_id, end_id, step, save_path=Non...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
benchmark/dataset_extract/dataset_extract_scannet.py
null
null
null
null
null
null
Python
2026-05-04T02:50:00.301853
import os import numpy as np import os.path as osp from PIL import Image from tqdm import tqdm import csv import cv2 import json import glob from natsort import natsorted import shutil from eval_utils import gen_json, gen_json_scannet_tae, get_sorted_files, copy_crop_files def extract_scannet( root, sample...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
benchmark/eval/eval_tae.py
null
null
null
null
null
null
Python
2026-05-04T02:50:00.306013
import numpy as np import cv2 import matplotlib.pyplot as plt import json import argparse from scipy.ndimage import map_coordinates from tqdm import tqdm import os import gc import time import torch device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') def compute_errors_torch(gt, pred): abs_rel =...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
benchmark/dataset_extract/dataset_extract_nyuv2.py
null
null
null
null
null
null
Python
2026-05-04T02:50:00.313030
import os import numpy as np import os.path as osp from PIL import Image from tqdm import tqdm import csv import cv2 import json import glob from natsort import natsorted import shutil from eval_utils import gen_json, get_sorted_files, copy_crop_files def extract_nyuv2( root, sample_len=-1, datatset_name=...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
app.py
null
null
null
null
null
null
Python
2026-05-04T02:50:00.314897
# Copyright (2025) Bytedance Ltd. and/or its affiliates # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # http://www.apache.org/licenses/LICENSE-2.0 # Unless required by applicable ...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
benchmark/eval/eval.py
null
null
null
null
null
null
Python
2026-05-04T02:50:00.346056
import numpy as np import cv2 import matplotlib.pyplot as plt import json import argparse from scipy.ndimage import map_coordinates from tqdm import tqdm import os import gc import torch from metric import * import metric device = 'cuda' eval_metrics = [ "abs_relative_difference", "rmse_linear", "delta1...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
loss/loss.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.127288
import torch import torch.nn as nn import numpy as np def reduction_batch_based(image_loss, M): # average of all valid pixels of the batch # avoid division by 0 (if sum(M) = sum(sum(mask)) = 0: sum(image_loss) = 0) divisor = torch.sum(M) if divisor == 0: return torch.sum(image_loss) * 0.0 ...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
run_streaming.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.128608
# Copyright (2025) Bytedance Ltd. and/or its affiliates # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # http://www.apache.org/licenses/LICENSE-2.0 # Unless required by applicable law o...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
run.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.151747
# Copyright (2025) Bytedance Ltd. and/or its affiliates # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # http://www.apache.org/licenses/LICENSE-2.0 # Unless required by applicable law o...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
benchmark/infer/infer.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.166160
import argparse import os import cv2 import json import torch from tqdm import tqdm import numpy as np from video_depth_anything.video_depth import VideoDepthAnything from utils.dc_utils import read_video_frames if __name__ == '__main__': parser = argparse.ArgumentParser() parser.add_argument('--infer_path', ...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
benchmark/eval/metric.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.229121
import torch def abs_relative_difference(output, target, valid_mask=None): actual_output = output actual_target = target abs_relative_diff = torch.abs(actual_output - actual_target) / actual_target if valid_mask is not None: abs_relative_diff[~valid_mask] = 0 n = valid_mask.sum((-1, -2)...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/dinov2_layers/drop_path.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.771462
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. # References: # https://github.com/facebookresearch/dino/blob/master/vision_transformer.py # https://github.com/rwigh...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/dinov2_layers/layer_scale.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.793757
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. # Modified from: https://github.com/huggingface/pytorch-image-models/blob/main/timm/models/vision_transformer.py#L103-L11...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/dinov2_layers/attention.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.795347
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. # References: # https://github.com/facebookresearch/dino/blob/master/vision_transformer.py # https://github.com/rwigh...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/dinov2_layers/block.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.796294
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. # References: # https://github.com/facebookresearch/dino/blob/master/vision_transformer.py # https://github.com/rwigh...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/dinov2_layers/mlp.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.811182
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. # References: # https://github.com/facebookresearch/dino/blob/master/vision_transformer.py # https://github.com/rwigh...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
utils/util.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.903485
# Copyright (2025) Bytedance Ltd. and/or its affiliates # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # http://www.apache.org/licenses/LICENSE-2.0 # Unless required by applicable ...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
loss/test_loss.py
null
null
null
null
null
null
Python
2026-05-04T02:50:01.988679
import torch from loss import VideoDepthLoss B = 2 T = 32 H = 518 W = 518 prediction = torch.randn(B, T, H, W).cuda() target = torch.randn(B, T, H, W).cuda() mask = torch.ones(B, T, H, W).bool().cuda() loss = VideoDepthLoss() val = loss(prediction, target, mask) print(f'loss: {val}')
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/dinov2.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.004596
# Copyright (c) Meta Platforms, Inc. and affiliates. # # This source code is licensed under the Apache License, Version 2.0 # found in the LICENSE file in the root directory of this source tree. # References: # https://github.com/facebookresearch/dino/blob/main/vision_transformer.py # https://github.com/rwightman/...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
utils/dc_utils.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.025465
# This file is originally from DepthCrafter/depthcrafter/utils.py at main · Tencent/DepthCrafter # SPDX-License-Identifier: MIT License license # # This file may have been modified by ByteDance Ltd. and/or its affiliates on [date of modification] # Original file is released under [ MIT License license], with the full l...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/dinov2_layers/__init__.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.028561
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. from .mlp import Mlp from .patch_embed import PatchEmbed from .swiglu_ffn import SwiGLUFFN, SwiGLUFFNFused from .block im...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/dinov2_layers/patch_embed.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.335609
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. # References: # https://github.com/facebookresearch/dino/blob/master/vision_transformer.py # https://github.com/rwigh...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/dpt_temporal.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.354449
# Copyright (2025) Bytedance Ltd. and/or its affiliates # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # http://www.apache.org/licenses/LICENSE-2.0 # Unless required by applicable ...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/dinov2_layers/swiglu_ffn.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.359738
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, Optional from torch import Tensor, nn import torch.nn.functional as F class SwiGLUFFN(nn....
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/dpt.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.375651
# Copyright (2025) Bytedance Ltd. and/or its affiliates # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # http://www.apache.org/licenses/LICENSE-2.0 # Unless required by applicable ...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/motion_module/attention.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.379372
# Copyright 2022 The HuggingFace Team. All rights reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicabl...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/motion_module/motion_module.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.433000
# This file is originally from AnimateDiff/animatediff/models/motion_module.py at main · guoyww/AnimateDiff # SPDX-License-Identifier: Apache-2.0 license # # This file may have been modified by ByteDance Ltd. and/or its affiliates on [date of modification] # Original file was released under [ Apache-2.0 license], with ...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/util/blocks.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.538165
import torch.nn as nn def _make_scratch(in_shape, out_shape, groups=1, expand=False): scratch = nn.Module() out_shape1 = out_shape out_shape2 = out_shape out_shape3 = out_shape if len(in_shape) >= 4: out_shape4 = out_shape if expand: out_shape1 = out_shape out_shape2 ...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/util/transform.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.550932
import numpy as np import cv2 class Resize(object): """Resize sample to given size (width, height). """ def __init__( self, width, height, resize_target=True, keep_aspect_ratio=False, ensure_multiple_of=1, resize_method="lower_bound", image_...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/video_depth.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.571473
# Copyright (2025) Bytedance Ltd. and/or its affiliates # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # http://www.apache.org/licenses/LICENSE-2.0 # Unless required by applicable law o...
DepthAnything/Video-Depth-Anything
https://github.com/DepthAnything/Video-Depth-Anything
null
null
null
null
1,890
null
null
apache-2.0
null
null
null
null
null
null
null
video_depth_anything/video_depth_stream.py
null
null
null
null
null
null
Python
2026-05-04T02:50:02.601226
# Copyright (2025) Bytedance Ltd. and/or its affiliates # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # http://www.apache.org/licenses/LICENSE-2.0 # Unless required by applicable ...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
docs/examples/generate_wan_dataset.py
null
null
null
null
null
null
Python
2026-05-04T02:50:07.641171
import argparse import csv import os import random import pandas as pd import torch # --- Fixed Architectural Constants (Internal Only) --- # channel number of video lantent FIXED_C = 16 # sequence length of text FIXED_L_TEXT = 512 # embedding dimension of text FIXED_D_TEXT = 4096 # dimension of clip feature (video...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
docs/conf.py
null
null
null
null
null
null
Python
2026-05-04T02:50:07.707608
import os import sys sys.path.insert(0, os.path.abspath("../..")) # -- Project information ----------------------------------------------------- version_file = "../veomni/__init__.py" with open(version_file, encoding="utf-8") as f: try: version_line = next(line for line in f if line.startswith(...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
scripts/moe_ckpt_merge/moe_merge.py
null
null
null
null
null
null
Python
2026-05-04T02:50:07.708943
import os import shutil import warnings from argparse import ArgumentParser from dataclasses import dataclass from glob import glob from typing import Generator, List, Tuple import torch from safetensors.torch import safe_open from tqdm import tqdm from transformers import AutoConfig from veomni.models import save_mo...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
scripts/deepseek_v3/fp8_cast_bf16.py
null
null
null
null
null
null
Python
2026-05-04T02:50:07.742150
# adapted from https://huggingface.co/deepseek-ai/DeepSeek-V3/blob/main/inference/fp8_cast_bf16.py import json import os from argparse import ArgumentParser from glob import glob import torch from kernel import weight_dequant from safetensors.torch import load_file from tqdm import tqdm from veomni.models import save...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
scripts/ci/check_doc_task_paths.py
null
null
null
null
null
null
Python
2026-05-04T02:50:07.743225
#!/usr/bin/env python3 """Verify that task script paths referenced in docs shell blocks exist on disk.""" from __future__ import annotations import argparse import re import sys from pathlib import Path # Fenced code blocks used for copy-paste training commands SHELL_FENCE = re.compile(r"```(?:shell|bash|sh)\s*\n(....
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
scripts/deepseek_v3/kernel.py
null
null
null
null
null
null
Python
2026-05-04T02:50:07.825584
# copied from https://huggingface.co/deepseek-ai/DeepSeek-V3/blob/main/inference/kernel.py from typing import Tuple import torch import triton import triton.language as tl from triton import Config @triton.jit def act_quant_kernel(x_ptr, y_ptr, s_ptr, BLOCK_SIZE: tl.constexpr): pid = tl.program_id(axis=0) of...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
scripts/download_hf_model.py
null
null
null
null
null
null
Python
2026-05-04T02:50:07.832608
import argparse import os from huggingface_hub import snapshot_download """ python3 scripts/download_hf_model.py --repo_id deepseek-ai/Janus-1.3B --local_dir Janus-1.3B """ if __name__ == "__main__": parser = argparse.ArgumentParser() parser.add_argument("--repo_id", type=str, default="deepseek-ai/Janus-1.3...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
scripts/multimodal/convert_data/tom-and-jerry.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.242575
# pip install soundfile, librosa import argparse import math import os from datasets import Dataset from veomni.data.multimodal.video_utils import ( load_video_bytes_from_path, ) def load_dataset(dataset_path: str): captions_file = os.path.join(dataset_path, "captions.txt") videos_file = os.path.join(da...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
scripts/profile/merge_chrome_trace.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.284384
import argparse import glob import gzip import json import os from typing import List def merge_traces_direct(input_patterns: List[str], output: str): """ Directly merge PyTorch trace files without timeline alignment. Each rank will maintain its original timestamps. """ merged = {"schemaVersion": ...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
scripts/trim_safetensor_layers.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.290095
""" This script trims hidden layer number for a given safetensor dir so that we can test weight loading against large models like deepseek conveniently Example usage: python scripts/trim_safetensor_layers.py \ --model_dir /mnt/hdfs/tianle.zhong/models/unsloth-deepseek-v3.1-bf16-merged \ --out_dir /mnt/hdfs/tianle...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
scripts/verify_torch29_conv3d_issue.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.327738
# Copied from https://github.com/pytorch/pytorch/issues/166643#issue-3571291598 # Check it to see if using the suggested cudnn version with torch 2.9 has fixed the conv3d memory issue # # Test results (torch 2.9.1+cu129) on A100: # # With nvidia-cudnn-cu12==9.15.1.9 (forced version): # Peak memory before: 1.2622 GB #...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/deprecated_task/train_qwen_vl.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.389827
import json import os import time from dataclasses import asdict, dataclass, field from typing import Any, Dict, List, Optional import torch import torch.distributed as dist import wandb from torch.utils.checkpoint import set_checkpoint_debug_enabled from tqdm import trange from veomni.arguments import DataArguments,...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/deprecated_task/train_flux.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.407863
import os import time from dataclasses import dataclass, field from typing import Any, Dict, List, Literal, Optional import torch import torch.distributed as dist import wandb from tqdm import trange from transformers import ( AutoConfig, CLIPTokenizer, T5TokenizerFast, ) from veomni.arguments import Data...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/deprecated_task/train_torch.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.432778
import json import os import time from dataclasses import asdict from datetime import timedelta from typing import Any, Dict, List import torch import torch.distributed as dist import wandb from torch.utils.checkpoint import set_checkpoint_debug_enabled from tqdm import trange from veomni.arguments import VeOmniArgum...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/infer/infer_omni_model.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.820745
import json import os from dataclasses import asdict, dataclass, field import requests import torch from PIL import Image from veomni.arguments import InferArguments, parse_args from veomni.data import build_multimodal_chat_template from veomni.data.multimodal.multimodal_transform import mask_input_ids from veomni.mo...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/deprecated_task/train_wan.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.822714
import os import time from dataclasses import dataclass, field from typing import Any, Dict, List import torch import torch.distributed as dist import torch.nn.functional as F import wandb from tqdm import trange from veomni.arguments import ( DataArguments, ModelArguments, TrainingArguments, parse_ar...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/infer/infer_text.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.926304
import json import readline # noqa: F401 from dataclasses import asdict, dataclass, field import torch from transformers import AutoTokenizer, TextStreamer from veomni.arguments import InferArguments, parse_args from veomni.models import build_foundation_model from veomni.utils import helper logger = helper.create...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/infer/infer_qwen2_vl.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.945567
import json from dataclasses import asdict, dataclass, field import requests from PIL import Image from veomni.arguments import InferArguments, parse_args from veomni.models import build_foundation_model, build_processor from veomni.utils import helper from veomni.utils.device import get_device_type logger = helper...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/omni/train_omni_model.py
null
null
null
null
null
null
Python
2026-05-04T02:50:08.968274
import json import os import time from collections import defaultdict from dataclasses import asdict, dataclass, field from functools import partial from typing import Any, Dict, List, Optional import torch import wandb from torch import distributed as dist from tqdm import trange from veomni.arguments import DataArg...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/train_text.py
null
null
null
null
null
null
Python
2026-05-04T02:50:09.029760
from veomni.arguments import parse_args from veomni.trainer.text_trainer import TextTrainer, VeOmniArguments if __name__ == "__main__": args = parse_args(VeOmniArguments) trainer = TextTrainer(args) trainer.train()
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/train_dit.py
null
null
null
null
null
null
Python
2026-05-04T02:50:09.031056
from veomni.arguments import parse_args from veomni.trainer.dit_trainer import DiTTrainer, VeOmniDiTArguments if __name__ == "__main__": args = parse_args(VeOmniDiTArguments) trainer = DiTTrainer(args) trainer.train()
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/train_text_rl.py
null
null
null
null
null
null
Python
2026-05-04T02:50:09.387299
from veomni.arguments import parse_args from veomni.trainer.base_rl_trainer import BaseRLTrainer from veomni.trainer.text_trainer import TextTrainer, VeOmniArguments class TextRLTrainer(TextTrainer): base: BaseRLTrainer def __init__(self, args: VeOmniArguments) -> None: # BaseRLTrainer.__init__ is NO...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/train_text_dpo.py
null
null
null
null
null
null
Python
2026-05-04T02:50:09.425586
from veomni.arguments import parse_args from veomni.trainer.text_dpo_trainer import TextDPOTrainer, VeOmniDPOArguments if __name__ == "__main__": args = parse_args(VeOmniDPOArguments) trainer = TextDPOTrainer(args) trainer.train()
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/train_vlm.py
null
null
null
null
null
null
Python
2026-05-04T02:50:09.517654
from veomni.arguments import parse_args from veomni.trainer.vlm_trainer import VeOmniVLMArguments, VLMTrainer if __name__ == "__main__": args = parse_args(VeOmniVLMArguments) trainer = VLMTrainer(args) trainer.train()
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tasks/train_vlm_rl.py
null
null
null
null
null
null
Python
2026-05-04T02:50:09.518710
from veomni.arguments import parse_args from veomni.trainer.base_rl_trainer import BaseRLTrainer from veomni.trainer.vlm_trainer import VeOmniVLMArguments, VLMTrainer class VLMRLTrainer(VLMTrainer): base: BaseRLTrainer def __init__(self, args: VeOmniVLMArguments): # BaseRLTrainer.__init__ is NOT call...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
tests/checkpoints/checkpoint_verification_utils.py
null
null
null
null
null
null
Python
2026-05-04T02:50:09.623409
#!/usr/bin/env python # Copyright 2025 Bytedance Ltd. and/or its affiliates # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless requi...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
scripts/download_hf_data.py
null
null
null
null
null
null
Python
2026-05-04T02:50:09.922371
import argparse from huggingface_hub import snapshot_download """ python3 scripts/download_hf_data.py --repo_id HuggingFaceFW/fineweb --local_dir ./fineweb/ --allow_patterns sample/10BT/* """ if __name__ == "__main__": parser = argparse.ArgumentParser() parser.add_argument("--repo_id", type=str, default="H...
ByteDance-Seed/VeOmni
https://github.com/ByteDance-Seed/VeOmni
null
null
null
null
1,889
null
null
apache-2.0
null
null
null
null
null
null
null
scripts/moe_ckpt_merge/moe_split.py
null
null
null
null
null
null
Python
2026-05-04T02:50:09.927479
""" Reverse process of moe_merge.py - splits merged MoE expert weights back to individual experts. This script takes a HF checkpoint with stacked/fused expert weights and splits them back to the original per-expert format expected by HuggingFace safetensors. Supported input formats: - v4 veomni format (from moe_mer...