Chengfang Lu
e72bc061c5
Major changes: - New frontend (platform/web/): Vite + React 18 + TypeScript + Tailwind - 4-module navigation: 数据送标 / 模型管理 / 车队管理 / 系统管理 - Data catalog with charts (DMS/ADAS/Lane 3-tab view) - Quality review workflow (标注质检): Good/Fine/Bad scoring with auto-advance - Audit enhancements: batch operations, rejection categories, Feishu notifications - Operation audit log (操作日志) - World model simulation studio (仿真工坊) - Dataset version management with snapshots and diff - ADAS 7-class dataset integration (138K images organized + compressed) - User management with Feishu integration and pagination - CRUD/search/filter on all pages, card layout redesign - PIL-optimized image overlay rendering - Auto-snapshot on build, in_review workflow stage - Removed embedded algorithm code (now in workspace)
47 lines
2.3 KiB
Python
47 lines
2.3 KiB
Python
# Define a series of shortcuts for commandline use of main_*.py
|
|
# a_b equals --a-b in commandline
|
|
|
|
SHORTCUTS = dict(
|
|
exp_name=dict(keys=['train.exp_name', 'test.exp_name'], type=str,
|
|
help='Name of experiment'),
|
|
checkpoint=dict(keys=['train.checkpoint', 'test.checkpoint'], type=str,
|
|
help='Continue/Load from a previous checkpoint'),
|
|
device=dict(keys=['train.device', 'test.device'], type=str,
|
|
help='CPU is not recommended!'),
|
|
workers=dict(keys=['train.workers', 'test.workers'], type=int,
|
|
help='Number of workers (threads) when loading data.'
|
|
'Recommend value for training=~ batch size'),
|
|
batch_size=dict(keys=['train.batch_size', 'test.batch_size'], type=int,
|
|
help='input batch size. Recommend 4 times the training batch size in testing'),
|
|
save_dir=dict(keys=['train.save_dir', 'test.save_dir'], type=str,
|
|
help='Path prefix to save all files excluding tensorboard log.'),
|
|
|
|
val_num_steps=dict(keys=['train.val_num_steps'], type=int,
|
|
help='Validation frequency'),
|
|
world_size=dict(keys=['train.world_size'], type=int,
|
|
help='Number of distributed processes'),
|
|
dist_url=dict(keys=['train.dist_url'], type=str,
|
|
help='url used to set up distributed training'),
|
|
|
|
thresh=dict(keys=['test.thresh'], type=float,
|
|
help='Threshold for detection tasks.'),
|
|
|
|
lr=dict(keys=['optimizer.lr'], type=float,
|
|
help='Learning rate'),
|
|
weight_decay=dict(keys=['optimizer.weight_decay'], type=float,
|
|
help='Weight decay'),
|
|
|
|
warmup_steps=dict(keys=['lr_scheduler.warmup_steps'], type=int,
|
|
help='Learning rate warmup steps.'),
|
|
epochs=dict(keys=['lr_scheduler.epochs', 'train.num_epochs'], type=int,
|
|
help='Number of epochs')
|
|
)
|
|
|
|
DEPRECATION_MAP = dict(
|
|
continue_from=dict(valid='checkpoint', message=''),
|
|
do_not_save=dict(valid=None, message='Please delete the .pt files yourself!'),
|
|
method=dict(valid=None, message='Please use the config files to define models!'),
|
|
model=dict(valid=None, message='Please use the config files to define models!'),
|
|
backbone=dict(valid=None, message='Please use the config files to define models!')
|
|
)
|