import

argparse
import math
import os
import random
import sys
import time
from copy import deepcopy
from datetime import datetime
from pathlib import Path

import numpy as np
import torch
import [Link] as dist
import [Link] as nn
import yaml
from [Link] import DistributedDataParallel as DDP
from [Link] import SGD, Adam, AdamW, lr_scheduler
from tqdm import tqdm

FILE = Path(__file__).resolve()
ROOT = [Link][0] # YOLOv5 root directory
if str(ROOT) not in [Link]:
[Link](str(ROOT)) # add ROOT to PATH
ROOT = Path([Link](ROOT, [Link]())) # relative

import val # for end-of-epoch mAP

from [Link] import attempt_load
from [Link] import Model
from [Link] import check_anchors
from [Link] import check_train_batch_size
from [Link] import Callbacks
from [Link] import create_dataloader
from [Link] import attempt_download
from [Link] import (LOGGER, check_amp, check_dataset, check_file,
check_git_status, check_img_size,
check_requirements, check_suffix, check_version,
check_yaml, colorstr, get_latest_run,
increment_path, init_seeds, intersect_dicts,
labels_to_class_weights,
labels_to_image_weights, methods, one_cycle,
print_args, print_mutation, strip_optimizer)
from [Link] import Loggers
from [Link].wandb_utils import check_wandb_resume
from [Link] import ComputeLoss
from [Link] import fitness
from [Link] import plot_evolve, plot_labels
from utils.torch_utils import EarlyStopping, ModelEMA, de_parallel,
select_device, torch_distributed_zero_first

LOCAL_RANK = int([Link]('LOCAL_RANK', -1)) #

[Link]
RANK = int([Link]('RANK', -1))
WORLD_SIZE = int([Link]('WORLD_SIZE', 1))

def train(hyp, opt, device, callbacks): # hyp is path/to/[Link] or hyp

dictionary
save_dir, epochs, batch_size, weights, single_cls, evolve, data, cfg,
resume, noval, nosave, workers, freeze = \
Path(opt.save_dir), [Link], opt.batch_size, [Link],
opt.single_cls, [Link], [Link], [Link], \
[Link], [Link], [Link], [Link], [Link]
[Link]('on_pretrain_routine_start')

# Directories
w = save_dir / 'weights' # weights dir
([Link] if evolve else w).mkdir(parents=True, exist_ok=True) # make
dir
last, best = w / '[Link]', w / '[Link]'

# Hyperparameters
if isinstance(hyp, str):
with open(hyp, errors='ignore') as f:
hyp = yaml.safe_load(f) # load hyps dict
[Link](colorstr('hyperparameters: ') + ', '.join(f'{k}={v}' for k, v
in [Link]()))

# Save run settings

if not evolve:
with open(save_dir / '[Link]', 'w') as f:
yaml.safe_dump(hyp, f, sort_keys=False)
with open(save_dir / '[Link]', 'w') as f:
 yaml.safe_dump(vars(opt), f, sort_keys=False)

# Loggers
data_dict = None
if RANK in {-1, 0}:
loggers = Loggers(save_dir, weights, opt, hyp, LOGGER) # loggers
instance
if [Link]:
data_dict = [Link].data_dict
if resume:
weights, epochs, hyp, batch_size = [Link], [Link],
[Link], opt.batch_size

# Register actions
for k in methods(loggers):
callbacks.register_action(k, callback=getattr(loggers, k))

# Config
plots = not evolve and not [Link] # create plots
cuda = [Link] != 'cpu'
init_seeds(1 + RANK)
with torch_distributed_zero_first(LOCAL_RANK):
data_dict = data_dict or check_dataset(data) # check if None
train_path, val_path = data_dict['train'], data_dict['val']
nc = 1 if single_cls else int(data_dict['nc']) # number of classes
names = ['item'] if single_cls and len(data_dict['names']) != 1 else
data_dict['names'] # class names
assert len(names) == nc, f'{len(names)} names found for nc={nc} dataset
in {data}' # check
is_coco = isinstance(val_path, str) and
val_path.endswith('coco/[Link]') # COCO dataset

# Model
check_suffix(weights, '.pt') # check weights
pretrained = [Link]('.pt')
if pretrained:
with torch_distributed_zero_first(LOCAL_RANK):
weights = attempt_download(weights) # download if not found
locally
ckpt = [Link](weights, map_location='cpu') # load checkpoint to
CPU to avoid CUDA memory leak
model = Model(cfg or ckpt['model'].yaml, ch=3, nc=nc,
anchors=[Link]('anchors')).to(device) # create
exclude = ['anchor'] if (cfg or [Link]('anchors')) and not resume
else [] # exclude keys
csd = ckpt['model'].float().state_dict() # checkpoint state_dict as
FP32
csd = intersect_dicts(csd, model.state_dict(), exclude=exclude) #
intersect
model.load_state_dict(csd, strict=False) # load
[Link](f'Transferred {len(csd)}/{len(model.state_dict())} items
from {weights}') # report
else:
model = Model(cfg, ch=3, nc=nc,
anchors=[Link]('anchors')).to(device) # create
amp = check_amp(model) # check AMP

# Freeze
freeze = [f'model.{x}.' for x in (freeze if len(freeze) > 1 else
range(freeze[0]))] # layers to freeze
for k, v in model.named_parameters():
v.requires_grad = True # train all layers
if any(x in k for x in freeze):
[Link](f'freezing {k}')
v.requires_grad = False

# Image size
gs = max(int([Link]()), 32) # grid size (max stride)
imgsz = check_img_size([Link], gs, floor=gs * 2) # verify imgsz is
gs-multiple

# Batch size
if RANK == -1 and batch_size == -1: # single-GPU only, estimate best
batch size
batch_size = check_train_batch_size(model, imgsz, amp)
loggers.on_params_update({"batch_size": batch_size})

# Optimizer
nbs = 64 # nominal batch size
accumulate = max(round(nbs / batch_size), 1) # accumulate loss before
optimizing
hyp['weight_decay'] *= batch_size * accumulate / nbs # scale
weight_decay
[Link](f"Scaled weight_decay = {hyp['weight_decay']}")
 g = [], [], [] # optimizer parameter groups
bn = tuple(v for k, v in nn.__dict__.items() if 'Norm' in k) #
normalization layers, i.e. BatchNorm2d()
for v in [Link]():
if hasattr(v, 'bias') and isinstance([Link], [Link]): # bias
g[2].append([Link])
if isinstance(v, bn): # weight (no decay)
g[1].append([Link])
elif hasattr(v, 'weight') and isinstance([Link], [Link]): #
weight (with decay)
g[0].append([Link])

if [Link] == 'Adam':
optimizer = Adam(g[2], lr=hyp['lr0'], betas=(hyp['momentum'], 0.999))
# adjust beta1 to momentum
elif [Link] == 'AdamW':
optimizer = AdamW(g[2], lr=hyp['lr0'], betas=(hyp['momentum'],
0.999)) # adjust beta1 to momentum
else:
optimizer = SGD(g[2], lr=hyp['lr0'], momentum=hyp['momentum'],
nesterov=True)

optimizer.add_param_group({'params': g[0], 'weight_decay':

hyp['weight_decay']}) # add g0 with weight_decay
optimizer.add_param_group({'params': g[1]}) # add g1 (BatchNorm2d
weights)
[Link](f"{colorstr('optimizer:')} {type(optimizer).__name__} with
parameter groups "
f"{len(g[1])} weight (no decay), {len(g[0])} weight,
{len(g[2])} bias")
del g

# Scheduler
if opt.cos_lr:
lf = one_cycle(1, hyp['lrf'], epochs) # cosine 1->hyp['lrf']
else:
lf = lambda x: (1 - x / epochs) * (1.0 - hyp['lrf']) + hyp['lrf'] #
linear
scheduler = lr_scheduler.LambdaLR(optimizer, lr_lambda=lf) #
plot_lr_scheduler(optimizer, scheduler, epochs)
 # EMA
ema = ModelEMA(model) if RANK in {-1, 0} else None

# Resume
start_epoch, best_fitness = 0, 0.0
if pretrained:
# Optimizer
if ckpt['optimizer'] is not None:
optimizer.load_state_dict(ckpt['optimizer'])
best_fitness = ckpt['best_fitness']

# EMA
if ema and [Link]('ema'):
[Link].load_state_dict(ckpt['ema'].float().state_dict())
[Link] = ckpt['updates']

# Epochs
start_epoch = ckpt['epoch'] + 1
if resume:
assert start_epoch > 0, f'{weights} training to {epochs} epochs
is finished, nothing to resume.'
if epochs < start_epoch:
[Link](f"{weights} has been trained for {ckpt['epoch']}
epochs. Fine-tuning for {epochs} more epochs.")
epochs += ckpt['epoch'] # finetune additional epochs

del ckpt, csd

# DP mode
if cuda and RANK == -1 and [Link].device_count() > 1:
[Link]('WARNING: DP not recommended, use
[Link] for best DDP Multi-GPU results.\n'
'See Multi-GPU Tutorial at
[Link] to get started.')
model = [Link](model)

# SyncBatchNorm
if opt.sync_bn and cuda and RANK != -1:
model =
[Link].convert_sync_batchnorm(model).to(device)
[Link]('Using SyncBatchNorm()')

# Trainloader
train_loader, dataset = create_dataloader(train_path,
imgsz,
batch_size // WORLD_SIZE,
gs,
single_cls,
hyp=hyp,
augment=True,
cache=None if [Link] ==
'val' else [Link],
rect=[Link],
rank=LOCAL_RANK,
workers=workers,

image_weights=opt.image_weights,
quad=[Link],
prefix=colorstr('train: '),
shuffle=True)
mlc = int([Link]([Link], 0)[:, 0].max()) # max label
class
nb = len(train_loader) # number of batches
assert mlc < nc, f'Label class {mlc} exceeds nc={nc} in {data}. Possible
class labels are 0-{nc - 1}'

# Process 0
if RANK in {-1, 0}:
val_loader = create_dataloader(val_path,
imgsz,
batch_size // WORLD_SIZE * 2,
gs,
single_cls,
hyp=hyp,
cache=None if noval else [Link],
rect=True,
rank=-1,
workers=workers * 2,
pad=0.5,
prefix=colorstr('val: '))[0]

if not resume:
 labels = [Link]([Link], 0)
# c = [Link](labels[:, 0]) # classes
# cf = [Link]([Link](), minlength=nc) + 1. # frequency
# model._initialize_biases([Link](device))
if plots:
plot_labels(labels, names, save_dir)

# Anchors
if not [Link]:
check_anchors(dataset, model=model, thr=hyp['anchor_t'],
imgsz=imgsz)
[Link]().float() # pre-reduce anchor precision

[Link]('on_pretrain_routine_end')

# DDP mode
if cuda and RANK != -1:
if check_version(torch.__version__, '1.11.0'):
model = DDP(model, device_ids=[LOCAL_RANK],
output_device=LOCAL_RANK, static_graph=True)
else:
model = DDP(model, device_ids=[LOCAL_RANK],
output_device=LOCAL_RANK)

# Model attributes
nl = de_parallel(model).model[-1].nl # number of detection layers (to
scale hyps)
hyp['box'] *= 3 / nl # scale to layers
hyp['cls'] *= nc / 80 * 3 / nl # scale to classes and layers
hyp['obj'] *= (imgsz / 640) ** 2 * 3 / nl # scale to image size and
layers
hyp['label_smoothing'] = opt.label_smoothing
[Link] = nc # attach number of classes to model
[Link] = hyp # attach hyperparameters to model
model.class_weights = labels_to_class_weights([Link],
nc).to(device) * nc # attach class weights
[Link] = names

# Start training
t0 = [Link]()
nw = max(round(hyp['warmup_epochs'] * nb), 100) # number of warmup
iterations, max(3 epochs, 100 iterations)
# nw = min(nw, (epochs - start_epoch) / 2 * nb) # limit warmup to < 1/2
of training
last_opt_step = -1
maps = [Link](nc) # mAP per class
results = (0, 0, 0, 0, 0, 0, 0) # P, R, [email protected], [email protected],
val_loss(box, obj, cls)
scheduler.last_epoch = start_epoch - 1 # do not move
scaler = [Link](enabled=amp)
stopper = EarlyStopping(patience=[Link])
compute_loss = ComputeLoss(model) # init loss class
[Link]('on_train_start')
[Link](f'Image sizes {imgsz} train, {imgsz} val\n'
f'Using {train_loader.num_workers * WORLD_SIZE} dataloader
workers\n'
f"Logging results to {colorstr('bold', save_dir)}\n"
f'Starting training for {epochs} epochs...')
for epoch in range(start_epoch, epochs): # epoch
------------------------------------------------------------------
[Link]('on_train_epoch_start')
[Link]()

# Update image weights (optional, single-GPU only)

if opt.image_weights:
cw = model.class_weights.cpu().numpy() * (1 - maps) ** 2 / nc #
class weights
iw = labels_to_image_weights([Link], nc=nc,
class_weights=cw) # image weights
[Link] = [Link](range(dataset.n), weights=iw,
k=dataset.n) # rand weighted idx

# Update mosaic border (optional)

# b = int([Link](0.25 * imgsz, 0.75 * imgsz + gs) // gs * gs)
# dataset.mosaic_border = [b - imgsz, -b] # height, width borders

mloss = [Link](3, device=device) # mean losses

if RANK != -1:
train_loader.sampler.set_epoch(epoch)
pbar = enumerate(train_loader)
[Link](('\n' + '%10s' * 7) % ('Epoch', 'gpu_mem', 'box', 'obj',
'cls', 'labels', 'img_size'))
if RANK in {-1, 0}:
pbar = tqdm(pbar, total=nb, bar_format='{l_bar}{bar:10}{r_bar}
{bar:-10b}') # progress bar
optimizer.zero_grad()
for i, (imgs, targets, paths, _) in pbar: # batch
-------------------------------------------------------------
[Link]('on_train_batch_start')
ni = i + nb * epoch # number integrated batches (since train
start)
imgs = [Link](device, non_blocking=True).float() / 255 # uint8
to float32, 0-255 to 0.0-1.0

# Warmup
if ni <= nw:
xi = [0, nw] # x interp
# compute_loss.gr = [Link](ni, xi, [0.0, 1.0]) # iou loss
ratio (obj_loss = 1.0 or iou)
accumulate = max(1, [Link](ni, xi, [1, nbs /
batch_size]).round())
for j, x in enumerate(optimizer.param_groups):
# bias lr falls from 0.1 to lr0, all other lrs rise from
0.0 to lr0
x['lr'] = [Link](ni, xi, [hyp['warmup_bias_lr'] if j
== 2 else 0.0, x['initial_lr'] * lf(epoch)])
if 'momentum' in x:
x['momentum'] = [Link](ni, xi,
[hyp['warmup_momentum'], hyp['momentum']])

# Multi-scale
if opt.multi_scale:
sz = [Link](imgsz * 0.5, imgsz * 1.5 + gs) // gs *
gs # size
sf = sz / max([Link][2:]) # scale factor
if sf != 1:
ns = [[Link](x * sf / gs) * gs for x in
[Link][2:]] # new shape (stretched to gs-multiple)
imgs = [Link](imgs, size=ns,
mode='bilinear', align_corners=False)

# Forward
with [Link](amp):
pred = model(imgs) # forward
loss, loss_items = compute_loss(pred, [Link](device)) #
loss scaled by batch_size
if RANK != -1:
 loss *= WORLD_SIZE # gradient averaged between devices
in DDP mode
if [Link]:
loss *= 4.

# Backward
[Link](loss).backward()

# Optimize
if ni - last_opt_step >= accumulate:
[Link](optimizer) # [Link]
[Link]()
optimizer.zero_grad()
if ema:
[Link](model)
last_opt_step = ni

# Log
if RANK in {-1, 0}:
mloss = (mloss * i + loss_items) / (i + 1) # update mean
losses
mem = f'{[Link].memory_reserved() / 1E9 if
[Link].is_available() else 0:.3g}G' # (GB)
pbar.set_description(('%10s' * 2 + '%10.4g' * 5) %
(f'{epoch}/{epochs - 1}', mem, *mloss,
[Link][0], [Link][-1]))
[Link]('on_train_batch_end', ni, model, imgs, targets,
paths, plots)
if callbacks.stop_training:
return
# end batch
-----------------------------------------------------------------------------
-------------------

# Scheduler
lr = [x['lr'] for x in optimizer.param_groups] # for loggers
[Link]()

if RANK in {-1, 0}:

# mAP
[Link]('on_train_epoch_end', epoch=epoch)
 ema.update_attr(model, include=['yaml', 'nc', 'hyp', 'names',
'stride', 'class_weights'])
final_epoch = (epoch + 1 == epochs) or stopper.possible_stop
if not noval or final_epoch: # Calculate mAP
results, maps, _ = [Link](data_dict,
batch_size=batch_size //
WORLD_SIZE * 2,
imgsz=imgsz,
model=[Link],
single_cls=single_cls,
dataloader=val_loader,
save_dir=save_dir,
plots=False,
callbacks=callbacks,
compute_loss=compute_loss)

# Update best mAP

fi = fitness([Link](results).reshape(1, -1)) # weighted
combination of [P, R, [email protected], [email protected]]
if fi > best_fitness:
best_fitness = fi
log_vals = list(mloss) + list(results) + lr
[Link]('on_fit_epoch_end', log_vals, epoch, best_fitness,
fi)

# Save model
if (not nosave) or (final_epoch and not evolve): # if save
ckpt = {
'epoch': epoch,
'best_fitness': best_fitness,
'model': deepcopy(de_parallel(model)).half(),
'ema': deepcopy([Link]).half(),
'updates': [Link],
'optimizer': optimizer.state_dict(),
'wandb_id': [Link].wandb_run.id if [Link]
else None,
'date': [Link]().isoformat()}

# Save last, best and delete

[Link](ckpt, last)
if best_fitness == fi:
[Link](ckpt, best)
if opt.save_period > 0 and epoch % opt.save_period == 0:
 [Link](ckpt, w / f'epoch{epoch}.pt')
del ckpt
[Link]('on_model_save', last, epoch, final_epoch,
best_fitness, fi)

# Stop Single-GPU
if RANK == -1 and stopper(epoch=epoch, fitness=fi):
break

# Stop DDP TODO: known issues

s[Link]
# stop = stopper(epoch=epoch, fitness=fi)
# if RANK == 0:
# dist.broadcast_object_list([stop], 0) # broadcast 'stop' to
all ranks

# Stop DPP
# with torch_distributed_zero_first(RANK):
# if stop:
# break # must break all DDP ranks

# end epoch
-----------------------------------------------------------------------------
-----------------------
# end training
-----------------------------------------------------------------------------
------------------------
if RANK in {-1, 0}:
[Link](f'\n{epoch - start_epoch + 1} epochs completed in
{([Link]() - t0) / 3600:.3f} hours.')
for f in last, best:
if [Link]():
strip_optimizer(f) # strip optimizers
if f is best:
[Link](f'\nValidating {f}...')
results, _, _ = [Link](
data_dict,
batch_size=batch_size // WORLD_SIZE * 2,
imgsz=imgsz,
model=attempt_load(f, device).half(),
iou_thres=0.65 if is_coco else 0.60, # best
pycocotools results at 0.65
 single_cls=single_cls,
dataloader=val_loader,
save_dir=save_dir,
save_json=is_coco,
verbose=True,
plots=plots,
callbacks=callbacks,
compute_loss=compute_loss) # val best model with
plots
if is_coco:
[Link]('on_fit_epoch_end', list(mloss) +
list(results) + lr, epoch, best_fitness, fi)

[Link]('on_train_end', last, best, plots, epoch, results)

[Link].empty_cache()
return results

def parse_opt(known=False):
parser = [Link]()
parser.add_argument('--weights', type=str, default=ROOT / '[Link]',
help='initial weights path')
parser.add_argument('--cfg', type=str, default='', help='[Link]
path')
parser.add_argument('--data', type=str, default=ROOT /
'data/[Link]', help='[Link] path')
parser.add_argument('--hyp', type=str, default=ROOT /
'data/hyps/[Link]', help='hyperparameters path')
parser.add_argument('--epochs', type=int, default=300)
parser.add_argument('--batch-size', type=int, default=16, help='total
batch size for all GPUs, -1 for autobatch')
parser.add_argument('--imgsz', '--img', '--img-size', type=int,
default=640, help='train, val image size (pixels)')
parser.add_argument('--rect', action='store_true', help='rectangular
training')
parser.add_argument('--resume', nargs='?', const=True, default=False,
help='resume most recent training')
parser.add_argument('--nosave', action='store_true', help='only save
final checkpoint')
parser.add_argument('--noval', action='store_true', help='only validate
final epoch')
 parser.add_argument('--noautoanchor', action='store_true', help='disable
AutoAnchor')
parser.add_argument('--noplots', action='store_true', help='save no plot
files')
parser.add_argument('--evolve', type=int, nargs='?', const=300,
help='evolve hyperparameters for x generations')
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('--device', default='', help='cuda device, i.e. 0 or
0,1,2,3 or cpu')
parser.add_argument('--multi-scale', action='store_true', help='vary img-
size +/- 50%%')
parser.add_argument('--single-cls', action='store_true', help='train
multi-class data as single-class')
parser.add_argument('--optimizer', type=str, choices=['SGD', 'Adam',
'AdamW'], default='SGD', help='optimizer')
parser.add_argument('--sync-bn', action='store_true', help='use
SyncBatchNorm, only available in DDP mode')
parser.add_argument('--workers', type=int, default=8, help='max
dataloader workers (per RANK in DDP mode)')
parser.add_argument('--project', default=ROOT / 'runs/train', help='save
to project/name')
parser.add_argument('--name', default='exp', help='save to project/name')
parser.add_argument('--exist-ok', action='store_true', help='existing
project/name ok, do not increment')
parser.add_argument('--quad', action='store_true', help='quad
dataloader')
parser.add_argument('--cos-lr', action='store_true', help='cosine LR
scheduler')
parser.add_argument('--label-smoothing', type=float, default=0.0,
help='Label smoothing epsilon')
parser.add_argument('--patience', type=int, default=100,
help='EarlyStopping patience (epochs without improvement)')
parser.add_argument('--freeze', nargs='+', type=int, default=[0],
help='Freeze layers: backbone=10, first3=0 1 2')
parser.add_argument('--save-period', type=int, default=-1, help='Save
checkpoint every x epochs (disabled if < 1)')
parser.add_argument('--local_rank', type=int, default=-1, help='DDP
parameter, do not modify')

# Weights & Biases arguments

 parser.add_argument('--entity', default=None, help='W&B: Entity')
parser.add_argument('--upload_dataset', nargs='?', const=True,
default=False, help='W&B: Upload data, "val" option')
parser.add_argument('--bbox_interval', type=int, default=-1, help='W&B:
Set bounding-box image logging interval')
parser.add_argument('--artifact_alias', type=str, default='latest',
help='W&B: Version of dataset artifact to use')

opt = parser.parse_known_args()[0] if known else parser.parse_args()

return opt

def main(opt, callbacks=Callbacks()):

# Checks
if RANK in {-1, 0}:
print_args(vars(opt))
check_git_status()
check_requirements(exclude=['thop'])

# Resume
if [Link] and not check_wandb_resume(opt) and not [Link]: #
resume an interrupted run
ckpt = [Link] if isinstance([Link], str) else
get_latest_run() # specified or most recent path
assert [Link](ckpt), 'ERROR: --resume checkpoint does not
exist'
with open(Path(ckpt).[Link] / '[Link]', errors='ignore') as
f:
opt = [Link](**yaml.safe_load(f)) # replace
[Link], [Link], [Link] = '', ckpt, True # reinstate
[Link](f'Resuming training from {ckpt}')
else:
[Link], [Link], [Link], [Link], [Link] = \
check_file([Link]), check_yaml([Link]), check_yaml([Link]),
str([Link]), str([Link]) # checks
assert len([Link]) or len([Link]), 'either --cfg or --weights
must be specified'
if [Link]:
if [Link] == str(ROOT / 'runs/train'): # if default project
name, rename to runs/evolve
[Link] = str(ROOT / 'runs/evolve')
opt.exist_ok, [Link] = [Link], False # pass resume to
exist_ok and disable resume
if [Link] == 'cfg':
[Link] = Path([Link]).stem # use [Link] as name
opt.save_dir = str(increment_path(Path([Link]) / [Link],
exist_ok=opt.exist_ok))

# DDP mode
device = select_device([Link], batch_size=opt.batch_size)
if LOCAL_RANK != -1:
msg = 'is not compatible with YOLOv5 Multi-GPU DDP training'
assert not opt.image_weights, f'--image-weights {msg}'
assert not [Link], f'--evolve {msg}'
assert opt.batch_size != -1, f'AutoBatch with --batch-size -1 {msg},
please pass a valid --batch-size'
assert opt.batch_size % WORLD_SIZE == 0, f'--batch-size
{opt.batch_size} must be multiple of WORLD_SIZE'
assert [Link].device_count() > LOCAL_RANK, 'insufficient CUDA
devices for DDP command'
[Link].set_device(LOCAL_RANK)
device = [Link]('cuda', LOCAL_RANK)
dist.init_process_group(backend="nccl" if dist.is_nccl_available()
else "gloo")

# Train
if not [Link]:
train([Link], opt, device, callbacks)
if WORLD_SIZE > 1 and RANK == 0:
[Link]('Destroying process group... ')
dist.destroy_process_group()

# 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)

with open([Link], errors='ignore') as f:

hyp = yaml.safe_load(f) # load hyps dict
if 'anchors' not in hyp: # anchors commented in [Link]
hyp['anchors'] = 3
[Link], [Link], save_dir = True, True, Path(opt.save_dir) #
only val/save final epoch
# ei = [isinstance(x, (int, float)) for x in [Link]()] #
evolvable indices
evolve_yaml, evolve_csv = save_dir / 'hyp_evolve.yaml', save_dir /
'[Link]'
if [Link]:
[Link](f'gsutil cp gs://{[Link]}/[Link] {evolve_csv}')
# download [Link] if exists

for _ in range([Link]): # generations to evolve

 if evolve_csv.exists(): # if [Link] exists: select best hyps
and mutate
# Select parent(s)
parent = 'single' # parent selection method: 'single' or
'weighted'
x = [Link](evolve_csv, ndmin=2, delimiter=',',
skiprows=1)
n = min(5, len(x)) # number of previous results to consider
x = x[[Link](-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[[Link](0, n - 1)] # random selection
x = x[[Link](range(n), weights=w)[0]] # weighted
selection
elif parent == 'weighted':
x = (x * [Link](n, 1)).sum(0) / [Link]() # weighted
combination

# Mutate
mp, s = 0.8, 0.2 # mutation probability, sigma
npr = [Link]
[Link](int([Link]()))
g = [Link]([meta[k][0] for k in [Link]()]) # gains 0-1
ng = len(meta)
v = [Link](ng)
while all(v == 1): # mutate until a change occurs (prevent
duplicates)
v = (g * ([Link](ng) < mp) * [Link](ng) *
[Link]() * s + 1).clip(0.3, 3.0)
for i, k in enumerate([Link]()): # [Link]([Link](),
300)
hyp[k] = float(x[i + 7] * v[i]) # mutate

# Constrain to limits
for k, v in [Link]():
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([Link](), opt, device, callbacks)
callbacks = Callbacks()
# Write mutation results
 print_mutation(results, [Link](), save_dir, [Link])

# Plot results
plot_evolve(evolve_csv)
[Link](f'Hyperparameter evolution finished {[Link]}
generations\n'
f"Results saved to {colorstr('bold', save_dir)}\n"
f'Usage example: $ python [Link] --hyp {evolve_yaml}')

def run(**kwargs):
# Usage: import train; [Link](data='[Link]', imgsz=320,
weights='[Link]')
opt = parse_opt(True)
for k, v in [Link]():
setattr(opt, k, v)
main(opt)
return opt

if __name__ == "__main__":
opt = parse_opt()
main(opt)

Wild Animal Detection

import
argparse
import os
import sys
from pathlib import Path

import torch
import [Link] as cudnn

FILE = Path(__file__).resolve()
ROOT = [Link][0] # YOLOv5 root directory
if str(ROOT) not in [Link]:
[Link](str(ROOT)) # add ROOT to PATH
ROOT = Path([Link](ROOT, [Link]())) # relative

from [Link] import DetectMultiBackend

from [Link] import IMG_FORMATS, VID_FORMATS, LoadImages,
LoadStreams
from [Link] import (LOGGER, check_file, check_img_size, check_imshow,
check_requirements, colorstr, cv2,
increment_path, non_max_suppression, print_args,
scale_coords, strip_optimizer, xyxy2xywh)
from [Link] import Annotator, colors, save_one_box
from utils.torch_utils import select_device, time_sync

@torch.no_grad()
def run(
weights=ROOT / '[Link]', # [Link] path(s)
source=ROOT / 'data/images', # file/dir/URL/glob, 0 for webcam
data=ROOT / 'data/[Link]', # [Link] path
imgsz=(640, 640), # inference size (height, width)
conf_thres=0.25, # confidence threshold
iou_thres=0.45, # NMS IOU threshold
max_det=1000, # maximum detections per image
device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu
view_img=False, # show results
save_txt=False, # save results to *.txt
save_conf=False, # save confidences in --save-txt labels
save_crop=False, # save cropped prediction boxes
nosave=False, # do not save images/videos
classes=15, # filter by class: --class 0, or --class 0 2 3
agnostic_nms=False, # class-agnostic NMS
augment=False, # augmented inference
visualize=False, # visualize features
update=False, # update all models
project=ROOT / 'runs/detect', # save results to project/name
name='exp', # save results to project/name
exist_ok=False, # existing project/name ok, do not increment
line_thickness=3, # bounding box thickness (pixels)
hide_labels=False, # hide labels
hide_conf=False, # hide confidences
half=False, # use FP16 half-precision inference
dnn=False, # use OpenCV DNN for ONNX inference
):
source = str(source)
save_img = not nosave and not [Link]('.txt') # save inference
images
is_file = Path(source).suffix[1:] in (IMG_FORMATS + VID_FORMATS)
is_url = [Link]().startswith(('rtsp://', 'rtmp://', '[Link]
'[Link]
webcam = [Link]() or [Link]('.txt') or (is_url and not
is_file)
if is_url and is_file:
source = check_file(source) # download

# Directories
save_dir = increment_path(Path(project) / name, exist_ok=exist_ok) #
increment run
(save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True,
exist_ok=True) # make dir

# Load model
device = select_device(device)
model = DetectMultiBackend(weights, device=device, dnn=dnn, data=data,
fp16=half)
stride, names, pt = [Link], [Link], [Link]
imgsz = check_img_size(imgsz, s=stride) # check image size

# Dataloader
if webcam:
view_img = check_imshow()
[Link] = True # set True to speed up constant image size
inference
dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt)
bs = len(dataset) # batch_size
else:
dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt)
bs = 1 # batch_size
vid_path, vid_writer = [None] * bs, [None] * bs

# Run inference
[Link](imgsz=(1 if pt else bs, 3, *imgsz)) # warmup
dt, seen = [0.0, 0.0, 0.0], 0
for path, im, im0s, vid_cap, s in dataset:
t1 = time_sync()
 im = torch.from_numpy(im).to(device)
im = [Link]() if model.fp16 else [Link]() # uint8 to fp16/32
im /= 255 # 0 - 255 to 0.0 - 1.0
if len([Link]) == 3:
im = im[None] # expand for batch dim
t2 = time_sync()
dt[0] += t2 - t1

# Inference
visualize = increment_path(save_dir / Path(path).stem, mkdir=True) if
visualize else False
pred = model(im, augment=augment, visualize=visualize)
t3 = time_sync()
dt[1] += t3 - t2

# NMS
pred = non_max_suppression(pred, conf_thres, iou_thres, classes,
agnostic_nms, max_det=max_det)
dt[2] += time_sync() - t3

# Second-stage classifier (optional)

# pred = [Link].apply_classifier(pred, classifier_model, im,
im0s)

# Process predictions
for i, det in enumerate(pred): # per image
seen += 1
if webcam: # batch_size >= 1
p, im0, frame = path[i], im0s[i].copy(), [Link]
s += f'{i}: '
else:
p, im0, frame = path, [Link](), getattr(dataset, 'frame',
0)

p = Path(p) # to Path
save_path = str(save_dir / [Link]) # [Link]
txt_path = str(save_dir / 'labels' / [Link]) + ('' if
[Link] == 'image' else f'_{frame}') # [Link]
s += '%gx%g ' % [Link][2:] # print string
gn = [Link]([Link])[[1, 0, 1, 0]] # normalization gain
whwh
 imc = [Link]() if save_crop else im0 # for save_crop
annotator = Annotator(im0, line_width=line_thickness,
example=str(names))
if len(det):
# Rescale boxes from img_size to im0 size
det[:, :4] = scale_coords([Link][2:], det[:, :4],
[Link]).round()

# Print results
for c in det[:, -1].unique():
n = (det[:, -1] == c).sum() # detections per class
s += f"{n} {names[int(c)]}{'s' * (n > 1)}, " # add to
string

# Write results
for *xyxy, conf, cls in reversed(det):
if save_txt: # Write to file
xywh = (xyxy2xywh([Link](xyxy).view(1, 4)) /
gn).view(-1).tolist() # normalized xywh
line = (cls, *xywh, conf) if save_conf else (cls,
*xywh) # label format
with open(f'{txt_path}.txt', 'a') as f:
[Link](('%g ' * len(line)).rstrip() % line + '\
n')

if save_img or save_crop or view_img: # Add bbox to

image
c = int(cls) # integer class
label = None if hide_labels else (names[c] if
hide_conf else f'{names[c]} {conf:.2f}')
annotator.box_label(xyxy, label, color=colors(c,
True))
if save_crop:
save_one_box(xyxy, imc, file=save_dir / 'crops' /
names[c] / f'{[Link]}.jpg', BGR=True)

# Stream results
im0 = [Link]()
if view_img:
[Link](str(p), im0)
[Link](1) # 1 millisecond
 # Save results (image with detections)
if save_img:
if [Link] == 'image':
[Link](save_path, im0)
else: # 'video' or 'stream'
if vid_path[i] != save_path: # new video
vid_path[i] = save_path
if isinstance(vid_writer[i], [Link]):
vid_writer[i].release() # release previous video
writer
if vid_cap: # video
fps = vid_cap.get(cv2.CAP_PROP_FPS)
w = int(vid_cap.get(cv2.CAP_PROP_FRAME_WIDTH))
h = int(vid_cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
else: # stream
fps, w, h = 30, [Link][1], [Link][0]
save_path = str(Path(save_path).with_suffix('.mp4'))
# force *.mp4 suffix on results videos
vid_writer[i] = [Link](save_path,
cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h))
vid_writer[i].write(im0)

# Print time (inference-only)

[Link](f'{s}Done. ({t3 - t2:.3f}s)')

# Print results
t = tuple(x / seen * 1E3 for x in dt) # speeds per image
[Link](f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per
image at shape {(1, 3, *imgsz)}' % t)
if save_txt or save_img:
s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to
{save_dir / 'labels'}" if save_txt else ''
[Link](f"Results saved to {colorstr('bold', save_dir)}{s}")
if update:
strip_optimizer(weights) # update model (to fix SourceChangeWarning)

def parse_opt():
parser = [Link]()
parser.add_argument('--weights', nargs='+', type=str, default=ROOT /
'[Link]', help='model path(s)')
 parser.add_argument('--source', type=str, default=ROOT / 'data/images',
help='file/dir/URL/glob, 0 for webcam')
parser.add_argument('--data', type=str, default=ROOT /
'data/[Link]', help='(optional) [Link] path')
parser.add_argument('--imgsz', '--img', '--img-size', nargs='+',
type=int, default=[640], help='inference size h,w')
parser.add_argument('--conf-thres', type=float, default=0.25,
help='confidence threshold')
parser.add_argument('--iou-thres', type=float, default=0.45, help='NMS
IoU threshold')
parser.add_argument('--max-det', type=int, default=1000, help='maximum
detections per image')
parser.add_argument('--device', default='', help='cuda device, i.e. 0 or
0,1,2,3 or cpu')
parser.add_argument('--view-img', action='store_true', help='show
results')
parser.add_argument('--save-txt', action='store_true', help='save results
to *.txt')
parser.add_argument('--save-conf', action='store_true', help='save
confidences in --save-txt labels')
parser.add_argument('--save-crop', action='store_true', help='save
cropped prediction boxes')
parser.add_argument('--nosave', action='store_true', help='do not save
images/videos')
parser.add_argument('--classes', nargs='+', type=int, help='filter by
class: --classes 0, or --classes 0 2 3')
parser.add_argument('--agnostic-nms', action='store_true', help='class-
agnostic NMS')
parser.add_argument('--augment', action='store_true', help='augmented
inference')
parser.add_argument('--visualize', action='store_true', help='visualize
features')
parser.add_argument('--update', action='store_true', help='update all
models')
parser.add_argument('--project', default=ROOT / 'runs/detect', help='save
results to project/name')
parser.add_argument('--name', default='exp', help='save results to
project/name')
parser.add_argument('--exist-ok', action='store_true', help='existing
project/name ok, do not increment')
parser.add_argument('--line-thickness', default=3, type=int,
help='bounding box thickness (pixels)')
parser.add_argument('--hide-labels', default=False, action='store_true',
help='hide labels')
parser.add_argument('--hide-conf', default=False, action='store_true',
help='hide confidences')
 parser.add_argument('--half', action='store_true', help='use FP16 half-
precision inference')
parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN
for ONNX inference')
opt = parser.parse_args()
[Link] *= 2 if len([Link]) == 1 else 1 # expand
print_args(vars(opt))
return opt

def main(opt):
check_requirements(exclude=('tensorboard', 'thop'))
run(**vars(opt))

if __name__ == "__main__":
opt = parse_opt()
main(opt)