if __name__ == "__main__":
opt = parse_opt() ##part1
main(opt) ##part2
part1
def parse_opt(known=False):
parser = argparse.ArgumentParser()
# ?
parser.add_argument('--hyp', type=str, default='data/hyps/hyp.scratch.yaml', help='hyperparameters path')
parser.add_argument('--single-cls', action='store_true', help='train multi-class data as single-class')
# ***** Rectangular training/inference去除这些原图仿射变换到(640,640)的冗余信息
parser.add_argument('--rect', action='store_true', help='rectangular training')
# ***** 它可能允许在较低 --img 尺寸下进行更高 --img 尺寸训练的一些好处
parser.add_argument('--quad', action='store_true', help='quad dataloader')
# ***** one hot -> label-smoothing
parser.add_argument('--label-smoothing', type=float, default=0.0, help='Label smoothing epsilon')
# ***** 超参数进化
parser.add_argument('--evolve', type=int, nargs='?', const=300, help='evolve hyperparameters for x generations')
parser.add_argument('--linear-lr', action='store_true', help='linear LR')
# resume参数
parser.add_argument('--resume', nargs='?', const=True, default=False, help='resume most recent training')
# 默认就行
parser.add_argument('--bucket', type=str, default='', help='gsutil bucket')
parser.add_argument('--cache', type=str, nargs='?', const='ram', help='--cache images in "ram" (default) or "disk"')
parser.add_argument('--image-weights', action='store_true', help='use weighted image selection for training')
parser.add_argument('--multi-scale', default=1, help='vary img-size +/- 50%%')
parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment')
parser.add_argument('--bbox_interval', type=int, default=-1, help='Set bounding-box image logging interval for W&B')
parser.add_argument('--save_period', type=int, default=-1, help='Log model after every "save_period" epoch')
parser.add_argument('--artifact_alias', type=str, default="latest", help='version of dataset artifact to be used')
parser.add_argument('--sync-bn', action='store_true', help='use SyncBatchNorm, only available in DDP mode')
#parser.add_argument('--upload_dataset', action='store_true', help='Upload dataset as W&B artifact table')
opt = parser.parse_known_args()[0] if known else parser.parse_args()
opt.nosave = opt.noval = opt.noautoanchor = False # nothing
opt.device = opt.workers = opt.freeze = opt.adam = 0
opt.weights = 'runs/train/exp4/weights/best.pt'
# opt.cfg = 'models/yolov5s.yaml'
opt.cfg = ''
opt.data = r'data\nc4dataset.yaml'
## 设置 cfg 从头开始 设置weights 迁移学习 data里面将yaml的nc改为需要的nc
opt.epochs = 500
opt.batch_size = 4
opt.imgsz = 640
opt.name = 'exp'
opt.project = 'runs/train'
opt.change_nc = False
opt.restart = False
return opt
part2
LOGGER = logging.getLogger(__name__)
LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1))
RANK = int(os.getenv('RANK', -1))
WORLD_SIZE = int(os.getenv('WORLD_SIZE', 1))
------------------------------------------------------------------------------------------------
local_rank和rank是一个意思,即代表第几个进程,world_size表示总共有n个进程
比如有2块gpu ,world_size = 5 , rank = 3,local_rank = 0 表示总共5个进程第 3 个进程内的第 1 块 GPU(不一定是0号gpu)。
local_rank和rank的取值范围是从0到n-1
------------------------------------------------------------------------------------------------
def main(opt):
set_logging(RANK)
------------------------------------------------------------------------------------------------
> def set_logging(rank=-1, verbose=True):
> logging.basicConfig(
> format="%(message)s",
> level=logging.INFO if (verbose and rank in [-1, 0]) else logging.WARN)
设置不同得日志等级,不用print出所有信息
------------------------------------------------------------------------------------------------
# Resume
if opt.resume and not check_wandb_resume(opt) and not opt.evolve: # resume an interrupted run
ckpt = opt.resume if isinstance(opt.resume, str) else get_latest_run() # specified or most recent path
assert os.path.isfile(ckpt), 'ERROR: --resume checkpoint does not exist'
with open(Path(ckpt).parent.parent / 'opt.yaml') as f:
opt = argparse.Namespace(**yaml.safe_load(f)) # replace
opt.cfg, opt.weights, opt.resume = '', ckpt, True # reinstate
LOGGER.info(f'Resuming training from {ckpt}')
else:
opt.data, opt.cfg, opt.hyp = check_file(opt.data), check_file(opt.cfg), check_file(opt.hyp) # check files
assert len(opt.cfg) or len(opt.weights), 'either --cfg or --weights must be specified'
if opt.evolve:
opt.project = 'runs/evolve'
opt.exist_ok = opt.resume
opt.save_dir = str(increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok))
------------------------------------------------------------------------------------------------
resume 在yolov5训练的pt模型中存储了ckpt['epoch'](好像是这个) 好像不太需要resume模块 目前还没用上
------------------------------------------------------------------------------------------------
# DDP mode 好像根本用不上
device = select_device(opt.device, batch_size=opt.batch_size)
# if LOCAL_RANK != -1:
# from datetime import timedelta
# assert torch.cuda.device_count() > LOCAL_RANK, 'insufficient CUDA devices for DDP command'
# assert opt.batch_size % WORLD_SIZE == 0, '--batch-size must be multiple of CUDA device count'
# assert not opt.image_weights, '--image-weights argument is not compatible with DDP training'
# assert not opt.evolve, '--evolve argument is not compatible with DDP training'
# assert not opt.sync_bn, '--sync-bn known training issue, see https://github.com/ultralytics/yolov5/issues/3998'
# torch.cuda.set_device(LOCAL_RANK)
# device = torch.device('cuda', LOCAL_RANK)
# dist.init_process_group(backend="nccl" if dist.is_nccl_available() else "gloo")
------------------------------------------------------------------------------------------------
DDP mode 多卡训练还是什么 目前我用不上
------------------------------------------------------------------------------------------------
# Train
if not opt.evolve:
train(opt.hyp, opt, device)
if WORLD_SIZE > 1 and RANK == 0:
_ = [print('Destroying process group... ', end=''), dist.destroy_process_group(), print('Done.')]
------------------------------------------------------------------------------------------------
如果不evlove 会将hyp opt device传入train
同时,train的所有预加载代码在这里结束了,下面全部是evolve相关代码
------------------------------------------------------------------------------------------------
# Evolve hyperparameters (optional)
else:
# Hyperparameter evolution metadata (mutation scale 0-1, lower_limit, upper_limit)
meta = {'lr0': (1, 1e-5, 1e-1), # initial learning rate (SGD=1E-2, Adam=1E-3)
'lrf': (1, 0.01, 1.0), # final OneCycleLR learning rate (lr0 * lrf)
'momentum': (0.3, 0.6, 0.98), # SGD momentum/Adam beta1
'weight_decay': (1, 0.0, 0.001), # optimizer weight decay
'warmup_epochs': (1, 0.0, 5.0), # warmup epochs (fractions ok)
'warmup_momentum': (1, 0.0, 0.95), # warmup initial momentum
'warmup_bias_lr': (1, 0.0, 0.2), # warmup initial bias lr
'box': (1, 0.02, 0.2), # box loss gain
'cls': (1, 0.2, 4.0), # cls loss gain
'cls_pw': (1, 0.5, 2.0), # cls BCELoss positive_weight
'obj': (1, 0.2, 4.0), # obj loss gain (scale with pixels)
'obj_pw': (1, 0.5, 2.0), # obj BCELoss positive_weight
'iou_t': (0, 0.1, 0.7), # IoU training threshold
'anchor_t': (1, 2.0, 8.0), # anchor-multiple threshold
'anchors': (2, 2.0, 10.0), # anchors per output grid (0 to ignore)
'fl_gamma': (0, 0.0, 2.0), # focal loss gamma (efficientDet default gamma=1.5)
'hsv_h': (1, 0.0, 0.1), # image HSV-Hue augmentation (fraction)
'hsv_s': (1, 0.0, 0.9), # image HSV-Saturation augmentation (fraction)
'hsv_v': (1, 0.0, 0.9), # image HSV-Value augmentation (fraction)
'degrees': (1, 0.0, 45.0), # image rotation (+/- deg)
'translate': (1, 0.0, 0.9), # image translation (+/- fraction)
'scale': (1, 0.0, 0.9), # image scale (+/- gain)
'shear': (1, 0.0, 10.0), # image shear (+/- deg)
'perspective': (0, 0.0, 0.001), # image perspective (+/- fraction), range 0-0.001
'flipud': (1, 0.0, 1.0), # image flip up-down (probability)
'fliplr': (0, 0.0, 1.0), # image flip left-right (probability)
'mosaic': (1, 0.0, 1.0), # image mixup (probability)
'mixup': (1, 0.0, 1.0), # image mixup (probability)
'copy_paste': (1, 0.0, 1.0)} # segment copy-paste (probability)
------------------------------------------------------------------------------------------------
如果evlove 会用上面的遗传算法边进化边训练(上面存储的应该是遗传算法的训练超参数)
------------------------------------------------------------------------------------------------
with open(opt.hyp) as f:
hyp = yaml.safe_load(f) # load hyps dict
if 'anchors' not in hyp: # anchors commented in hyp.yaml
hyp['anchors'] = 3
------------------------------------------------------------------------------------------------
这里用上了opt.hyp ,hyp中存储的是过往训练保留的一些可能有效的参数作为初始参数
------------------------------------------------------------------------------------------------
opt.noval, opt.nosave, save_dir = True, True, Path(opt.save_dir) # only val/save final epoch
# ei = [isinstance(x, (int, float)) for x in hyp.values()] # evolvable indices
evolve_yaml, evolve_csv = save_dir / 'hyp_evolve.yaml', save_dir / 'evolve.csv'
if opt.bucket:
os.system(f'gsutil cp gs://{opt.bucket}/evolve.csv {save_dir}') # download evolve.csv if exists
------------------------------------------------------------------------------------------------
不太理解这一句是在干什么
下面就是参数的更新,暂时用不上
------------------------------------------------------------------------------------------------
for _ in range(opt.evolve): # generations to evolve 迭代次数
if evolve_csv.exists(): # if evolve.csv exists: select best hyps and mutate
# Select parent(s)
parent = 'single' # parent selection method: 'single' or 'weighted'
x = np.loadtxt(evolve_csv, ndmin=2, delimiter=',', skiprows=1)
n = min(5, len(x)) # number of previous results to consider
x = x[np.argsort(-fitness(x))][:n] # top n mutations
w = fitness(x) - fitness(x).min() + 1E-6 # weights (sum > 0)
if parent == 'single' or len(x) == 1:
# x = x[random.randint(0, n - 1)] # random selection
x = x[random.choices(range(n), weights=w)[0]] # weighted selection
elif parent == 'weighted':
x = (x * w.reshape(n, 1)).sum(0) / w.sum() # weighted combination
# Mutate
mp, s = 0.8, 0.2 # mutation probability, sigma
npr = np.random
npr.seed(int(time.time()))
g = np.array([x[0] for x in meta.values()]) # gains 0-1
ng = len(meta)
v = np.ones(ng)
while all(v == 1): # mutate until a change occurs (prevent duplicates)
v = (g * (npr.random(ng) < mp) * npr.randn(ng) * npr.random() * s + 1).clip(0.3, 3.0)
for i, k in enumerate(hyp.keys()): # plt.hist(v.ravel(), 300)
hyp[k] = float(x[i + 7] * v[i]) # mutate
# Constrain to limits
for k, v in meta.items():
hyp[k] = max(hyp[k], v[1]) # lower limit
hyp[k] = min(hyp[k], v[2]) # upper limit
hyp[k] = round(hyp[k], 5) # significant digits
# Train mutation
results = train(hyp.copy(), opt, device)
# Write mutation results
print_mutation(results, hyp.copy(), save_dir, opt.bucket)
# Plot results
plot_evolve(evolve_csv)
print(f'Hyperparameter evolution finished\n'
f"Results saved to {colorstr('bold', save_dir)}\n"
f'Use best hyperparameters example: $ python train.py --hyp {evolve_yaml}')