Ви не можете вибрати більше 25 тем Теми мають розпочинатися з літери або цифри, можуть містити дефіси (-) і не повинні перевищувати 35 символів.

565 lines
29KB

  1. import argparse
  2. import logging
  3. import os
  4. import random
  5. import shutil
  6. import time
  7. from pathlib import Path
  8. from warnings import warn
  9. import math
  10. import numpy as np
  11. import torch.distributed as dist
  12. import torch.nn.functional as F
  13. import torch.optim as optim
  14. import torch.optim.lr_scheduler as lr_scheduler
  15. import torch.utils.data
  16. import yaml
  17. from torch.cuda import amp
  18. from torch.nn.parallel import DistributedDataParallel as DDP
  19. from torch.utils.tensorboard import SummaryWriter
  20. from tqdm import tqdm
  21. import test # import test.py to get mAP after each epoch
  22. from models.yolo import Model
  23. from utils.datasets import create_dataloader
  24. from utils.general import (
  25. torch_distributed_zero_first, labels_to_class_weights, plot_labels, check_anchors, labels_to_image_weights,
  26. compute_loss, plot_images, fitness, strip_optimizer, plot_results, get_latest_run, check_dataset, check_file,
  27. check_git_status, check_img_size, increment_dir, print_mutation, plot_evolution, set_logging, init_seeds)
  28. from utils.google_utils import attempt_download
  29. from utils.torch_utils import ModelEMA, select_device, intersect_dicts
  30. logger = logging.getLogger(__name__)
  31. def train(hyp, opt, device, tb_writer=None, wandb=None):
  32. logger.info(f'Hyperparameters {hyp}')
  33. log_dir = Path(tb_writer.log_dir) if tb_writer else Path(opt.logdir) / 'evolve' # logging directory
  34. wdir = log_dir / 'weights' # weights directory
  35. os.makedirs(wdir, exist_ok=True)
  36. last = wdir / 'last.pt'
  37. best = wdir / 'best.pt'
  38. results_file = str(log_dir / 'results.txt')
  39. epochs, batch_size, total_batch_size, weights, rank = \
  40. opt.epochs, opt.batch_size, opt.total_batch_size, opt.weights, opt.global_rank
  41. # Save run settings
  42. with open(log_dir / 'hyp.yaml', 'w') as f:
  43. yaml.dump(hyp, f, sort_keys=False)
  44. with open(log_dir / 'opt.yaml', 'w') as f:
  45. yaml.dump(vars(opt), f, sort_keys=False)
  46. # Configure
  47. cuda = device.type != 'cpu'
  48. init_seeds(2 + rank)
  49. with open(opt.data) as f:
  50. data_dict = yaml.load(f, Loader=yaml.FullLoader) # data dict
  51. with torch_distributed_zero_first(rank):
  52. check_dataset(data_dict) # check
  53. train_path = data_dict['train']
  54. test_path = data_dict['val']
  55. nc, names = (1, ['item']) if opt.single_cls else (int(data_dict['nc']), data_dict['names']) # number classes, names
  56. assert len(names) == nc, '%g names found for nc=%g dataset in %s' % (len(names), nc, opt.data) # check
  57. # Model
  58. pretrained = weights.endswith('.pt')
  59. if pretrained:
  60. with torch_distributed_zero_first(rank):
  61. attempt_download(weights) # download if not found locally
  62. ckpt = torch.load(weights, map_location=device) # load checkpoint
  63. if hyp.get('anchors'):
  64. ckpt['model'].yaml['anchors'] = round(hyp['anchors']) # force autoanchor
  65. model = Model(opt.cfg or ckpt['model'].yaml, ch=3, nc=nc).to(device) # create
  66. exclude = ['anchor'] if opt.cfg or hyp.get('anchors') else [] # exclude keys
  67. state_dict = ckpt['model'].float().state_dict() # to FP32
  68. state_dict = intersect_dicts(state_dict, model.state_dict(), exclude=exclude) # intersect
  69. model.load_state_dict(state_dict, strict=False) # load
  70. logger.info('Transferred %g/%g items from %s' % (len(state_dict), len(model.state_dict()), weights)) # report
  71. else:
  72. model = Model(opt.cfg, ch=3, nc=nc).to(device) # create
  73. # Freeze
  74. freeze = ['', ] # parameter names to freeze (full or partial)
  75. if any(freeze):
  76. for k, v in model.named_parameters():
  77. if any(x in k for x in freeze):
  78. print('freezing %s' % k)
  79. v.requires_grad = False
  80. # Optimizer
  81. nbs = 64 # nominal batch size
  82. accumulate = max(round(nbs / total_batch_size), 1) # accumulate loss before optimizing
  83. hyp['weight_decay'] *= total_batch_size * accumulate / nbs # scale weight_decay
  84. pg0, pg1, pg2 = [], [], [] # optimizer parameter groups
  85. for k, v in model.named_parameters():
  86. v.requires_grad = True
  87. if '.bias' in k:
  88. pg2.append(v) # biases
  89. elif '.weight' in k and '.bn' not in k:
  90. pg1.append(v) # apply weight decay
  91. else:
  92. pg0.append(v) # all else
  93. if opt.adam:
  94. optimizer = optim.Adam(pg0, lr=hyp['lr0'], betas=(hyp['momentum'], 0.999)) # adjust beta1 to momentum
  95. else:
  96. optimizer = optim.SGD(pg0, lr=hyp['lr0'], momentum=hyp['momentum'], nesterov=True)
  97. optimizer.add_param_group({'params': pg1, 'weight_decay': hyp['weight_decay']}) # add pg1 with weight_decay
  98. optimizer.add_param_group({'params': pg2}) # add pg2 (biases)
  99. logger.info('Optimizer groups: %g .bias, %g conv.weight, %g other' % (len(pg2), len(pg1), len(pg0)))
  100. del pg0, pg1, pg2
  101. # Scheduler https://arxiv.org/pdf/1812.01187.pdf
  102. # https://pytorch.org/docs/stable/_modules/torch/optim/lr_scheduler.html#OneCycleLR
  103. lf = lambda x: ((1 + math.cos(x * math.pi / epochs)) / 2) * (1 - hyp['lrf']) + hyp['lrf'] # cosine
  104. scheduler = lr_scheduler.LambdaLR(optimizer, lr_lambda=lf)
  105. # plot_lr_scheduler(optimizer, scheduler, epochs)
  106. # Logging
  107. if wandb and wandb.run is None:
  108. id = ckpt.get('wandb_id') if 'ckpt' in locals() else None
  109. wandb_run = wandb.init(config=opt, resume="allow", project="YOLOv5", name=os.path.basename(log_dir), id=id)
  110. # Resume
  111. start_epoch, best_fitness = 0, 0.0
  112. if pretrained:
  113. # Optimizer
  114. if ckpt['optimizer'] is not None:
  115. optimizer.load_state_dict(ckpt['optimizer'])
  116. best_fitness = ckpt['best_fitness']
  117. # Results
  118. if ckpt.get('training_results') is not None:
  119. with open(results_file, 'w') as file:
  120. file.write(ckpt['training_results']) # write results.txt
  121. # Epochs
  122. start_epoch = ckpt['epoch'] + 1
  123. if opt.resume:
  124. assert start_epoch > 0, '%s training to %g epochs is finished, nothing to resume.' % (weights, epochs)
  125. shutil.copytree(wdir, wdir.parent / f'weights_backup_epoch{start_epoch - 1}') # save previous weights
  126. if epochs < start_epoch:
  127. logger.info('%s has been trained for %g epochs. Fine-tuning for %g additional epochs.' %
  128. (weights, ckpt['epoch'], epochs))
  129. epochs += ckpt['epoch'] # finetune additional epochs
  130. del ckpt, state_dict
  131. # Image sizes
  132. gs = int(max(model.stride)) # grid size (max stride)
  133. imgsz, imgsz_test = [check_img_size(x, gs) for x in opt.img_size] # verify imgsz are gs-multiples
  134. # DP mode
  135. if cuda and rank == -1 and torch.cuda.device_count() > 1:
  136. model = torch.nn.DataParallel(model)
  137. # SyncBatchNorm
  138. if opt.sync_bn and cuda and rank != -1:
  139. model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(model).to(device)
  140. logger.info('Using SyncBatchNorm()')
  141. # Exponential moving average
  142. ema = ModelEMA(model) if rank in [-1, 0] else None
  143. # DDP mode
  144. if cuda and rank != -1:
  145. model = DDP(model, device_ids=[opt.local_rank], output_device=opt.local_rank)
  146. # Trainloader
  147. dataloader, dataset = create_dataloader(train_path, imgsz, batch_size, gs, opt,
  148. hyp=hyp, augment=True, cache=opt.cache_images, rect=opt.rect,
  149. rank=rank, world_size=opt.world_size, workers=opt.workers)
  150. mlc = np.concatenate(dataset.labels, 0)[:, 0].max() # max label class
  151. nb = len(dataloader) # number of batches
  152. assert mlc < nc, 'Label class %g exceeds nc=%g in %s. Possible class labels are 0-%g' % (mlc, nc, opt.data, nc - 1)
  153. # Process 0
  154. if rank in [-1, 0]:
  155. ema.updates = start_epoch * nb // accumulate # set EMA updates
  156. testloader = create_dataloader(test_path, imgsz_test, total_batch_size, gs, opt,
  157. hyp=hyp, augment=False, cache=opt.cache_images and not opt.notest, rect=True,
  158. rank=-1, world_size=opt.world_size, workers=opt.workers)[0] # testloader
  159. if not opt.resume:
  160. labels = np.concatenate(dataset.labels, 0)
  161. c = torch.tensor(labels[:, 0]) # classes
  162. # cf = torch.bincount(c.long(), minlength=nc) + 1. # frequency
  163. # model._initialize_biases(cf.to(device))
  164. plot_labels(labels, save_dir=log_dir)
  165. if tb_writer:
  166. # tb_writer.add_hparams(hyp, {}) # causes duplicate https://github.com/ultralytics/yolov5/pull/384
  167. tb_writer.add_histogram('classes', c, 0)
  168. # Anchors
  169. if not opt.noautoanchor:
  170. check_anchors(dataset, model=model, thr=hyp['anchor_t'], imgsz=imgsz)
  171. # Model parameters
  172. hyp['cls'] *= nc / 80. # scale coco-tuned hyp['cls'] to current dataset
  173. model.nc = nc # attach number of classes to model
  174. model.hyp = hyp # attach hyperparameters to model
  175. model.gr = 1.0 # iou loss ratio (obj_loss = 1.0 or iou)
  176. model.class_weights = labels_to_class_weights(dataset.labels, nc).to(device) # attach class weights
  177. model.names = names
  178. # Start training
  179. t0 = time.time()
  180. nw = max(round(hyp['warmup_epochs'] * nb), 1e3) # number of warmup iterations, max(3 epochs, 1k iterations)
  181. # nw = min(nw, (epochs - start_epoch) / 2 * nb) # limit warmup to < 1/2 of training
  182. maps = np.zeros(nc) # mAP per class
  183. results = (0, 0, 0, 0, 0, 0, 0) # P, R, mAP@.5, mAP@.5-.95, val_loss(box, obj, cls)
  184. scheduler.last_epoch = start_epoch - 1 # do not move
  185. scaler = amp.GradScaler(enabled=cuda)
  186. logger.info('Image sizes %g train, %g test\n'
  187. 'Using %g dataloader workers\nLogging results to %s\n'
  188. 'Starting training for %g epochs...' % (imgsz, imgsz_test, dataloader.num_workers, log_dir, epochs))
  189. for epoch in range(start_epoch, epochs): # epoch ------------------------------------------------------------------
  190. model.train()
  191. # Update image weights (optional)
  192. if opt.image_weights:
  193. # Generate indices
  194. if rank in [-1, 0]:
  195. cw = model.class_weights.cpu().numpy() * (1 - maps) ** 2 # class weights
  196. iw = labels_to_image_weights(dataset.labels, nc=nc, class_weights=cw) # image weights
  197. dataset.indices = random.choices(range(dataset.n), weights=iw, k=dataset.n) # rand weighted idx
  198. # Broadcast if DDP
  199. if rank != -1:
  200. indices = (torch.tensor(dataset.indices) if rank == 0 else torch.zeros(dataset.n)).int()
  201. dist.broadcast(indices, 0)
  202. if rank != 0:
  203. dataset.indices = indices.cpu().numpy()
  204. # Update mosaic border
  205. # b = int(random.uniform(0.25 * imgsz, 0.75 * imgsz + gs) // gs * gs)
  206. # dataset.mosaic_border = [b - imgsz, -b] # height, width borders
  207. mloss = torch.zeros(4, device=device) # mean losses
  208. if rank != -1:
  209. dataloader.sampler.set_epoch(epoch)
  210. pbar = enumerate(dataloader)
  211. logger.info(('\n' + '%10s' * 8) % ('Epoch', 'gpu_mem', 'box', 'obj', 'cls', 'total', 'targets', 'img_size'))
  212. if rank in [-1, 0]:
  213. pbar = tqdm(pbar, total=nb) # progress bar
  214. optimizer.zero_grad()
  215. for i, (imgs, targets, paths, _) in pbar: # batch -------------------------------------------------------------
  216. ni = i + nb * epoch # number integrated batches (since train start)
  217. imgs = imgs.to(device, non_blocking=True).float() / 255.0 # uint8 to float32, 0-255 to 0.0-1.0
  218. # Warmup
  219. if ni <= nw:
  220. xi = [0, nw] # x interp
  221. # model.gr = np.interp(ni, xi, [0.0, 1.0]) # iou loss ratio (obj_loss = 1.0 or iou)
  222. accumulate = max(1, np.interp(ni, xi, [1, nbs / total_batch_size]).round())
  223. for j, x in enumerate(optimizer.param_groups):
  224. # bias lr falls from 0.1 to lr0, all other lrs rise from 0.0 to lr0
  225. x['lr'] = np.interp(ni, xi, [hyp['warmup_bias_lr'] if j == 2 else 0.0, x['initial_lr'] * lf(epoch)])
  226. if 'momentum' in x:
  227. x['momentum'] = np.interp(ni, xi, [hyp['warmup_momentum'], hyp['momentum']])
  228. # Multi-scale
  229. if opt.multi_scale:
  230. sz = random.randrange(imgsz * 0.5, imgsz * 1.5 + gs) // gs * gs # size
  231. sf = sz / max(imgs.shape[2:]) # scale factor
  232. if sf != 1:
  233. ns = [math.ceil(x * sf / gs) * gs for x in imgs.shape[2:]] # new shape (stretched to gs-multiple)
  234. imgs = F.interpolate(imgs, size=ns, mode='bilinear', align_corners=False)
  235. # Forward
  236. with amp.autocast(enabled=cuda):
  237. pred = model(imgs) # forward
  238. loss, loss_items = compute_loss(pred, targets.to(device), model) # loss scaled by batch_size
  239. if rank != -1:
  240. loss *= opt.world_size # gradient averaged between devices in DDP mode
  241. # Backward
  242. scaler.scale(loss).backward()
  243. # Optimize
  244. if ni % accumulate == 0:
  245. scaler.step(optimizer) # optimizer.step
  246. scaler.update()
  247. optimizer.zero_grad()
  248. if ema:
  249. ema.update(model)
  250. # Print
  251. if rank in [-1, 0]:
  252. mloss = (mloss * i + loss_items) / (i + 1) # update mean losses
  253. mem = '%.3gG' % (torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0) # (GB)
  254. s = ('%10s' * 2 + '%10.4g' * 6) % (
  255. '%g/%g' % (epoch, epochs - 1), mem, *mloss, targets.shape[0], imgs.shape[-1])
  256. pbar.set_description(s)
  257. # Plot
  258. if ni < 3:
  259. f = str(log_dir / f'train_batch{ni}.jpg') # filename
  260. result = plot_images(images=imgs, targets=targets, paths=paths, fname=f)
  261. # if tb_writer and result is not None:
  262. # tb_writer.add_image(f, result, dataformats='HWC', global_step=epoch)
  263. # tb_writer.add_graph(model, imgs) # add model to tensorboard
  264. # end batch ------------------------------------------------------------------------------------------------
  265. # Scheduler
  266. lr = [x['lr'] for x in optimizer.param_groups] # for tensorboard
  267. scheduler.step()
  268. # DDP process 0 or single-GPU
  269. if rank in [-1, 0]:
  270. # mAP
  271. if ema:
  272. ema.update_attr(model, include=['yaml', 'nc', 'hyp', 'gr', 'names', 'stride'])
  273. final_epoch = epoch + 1 == epochs
  274. if not opt.notest or final_epoch: # Calculate mAP
  275. results, maps, times = test.test(opt.data,
  276. batch_size=total_batch_size,
  277. imgsz=imgsz_test,
  278. model=ema.ema,
  279. single_cls=opt.single_cls,
  280. dataloader=testloader,
  281. save_dir=log_dir,
  282. plots=epoch == 0 or final_epoch, # plot first and last
  283. log_imgs=opt.log_imgs)
  284. # Write
  285. with open(results_file, 'a') as f:
  286. f.write(s + '%10.4g' * 7 % results + '\n') # P, R, mAP@.5, mAP@.5-.95, val_loss(box, obj, cls)
  287. if len(opt.name) and opt.bucket:
  288. os.system('gsutil cp %s gs://%s/results/results%s.txt' % (results_file, opt.bucket, opt.name))
  289. # Log
  290. tags = ['train/giou_loss', 'train/obj_loss', 'train/cls_loss', # train loss
  291. 'metrics/precision', 'metrics/recall', 'metrics/mAP_0.5', 'metrics/mAP_0.5:0.95',
  292. 'val/giou_loss', 'val/obj_loss', 'val/cls_loss', # val loss
  293. 'x/lr0', 'x/lr1', 'x/lr2'] # params
  294. for x, tag in zip(list(mloss[:-1]) + list(results) + lr, tags):
  295. if tb_writer:
  296. tb_writer.add_scalar(tag, x, epoch) # tensorboard
  297. if wandb:
  298. wandb.log({tag: x}) # W&B
  299. # Update best mAP
  300. fi = fitness(np.array(results).reshape(1, -1)) # weighted combination of [P, R, mAP@.5, mAP@.5-.95]
  301. if fi > best_fitness:
  302. best_fitness = fi
  303. # Save model
  304. save = (not opt.nosave) or (final_epoch and not opt.evolve)
  305. if save:
  306. with open(results_file, 'r') as f: # create checkpoint
  307. ckpt = {'epoch': epoch,
  308. 'best_fitness': best_fitness,
  309. 'training_results': f.read(),
  310. 'model': ema.ema,
  311. 'optimizer': None if final_epoch else optimizer.state_dict(),
  312. 'wandb_id': wandb_run.id if wandb else None}
  313. # Save last, best and delete
  314. torch.save(ckpt, last)
  315. if best_fitness == fi:
  316. torch.save(ckpt, best)
  317. del ckpt
  318. # end epoch ----------------------------------------------------------------------------------------------------
  319. # end training
  320. if rank in [-1, 0]:
  321. # Strip optimizers
  322. n = opt.name if opt.name.isnumeric() else ''
  323. fresults, flast, fbest = log_dir / f'results{n}.txt', wdir / f'last{n}.pt', wdir / f'best{n}.pt'
  324. for f1, f2 in zip([wdir / 'last.pt', wdir / 'best.pt', results_file], [flast, fbest, fresults]):
  325. if os.path.exists(f1):
  326. os.rename(f1, f2) # rename
  327. if str(f2).endswith('.pt'): # is *.pt
  328. strip_optimizer(f2) # strip optimizer
  329. os.system('gsutil cp %s gs://%s/weights' % (f2, opt.bucket)) if opt.bucket else None # upload
  330. # Finish
  331. if not opt.evolve:
  332. plot_results(save_dir=log_dir) # save as results.png
  333. logger.info('%g epochs completed in %.3f hours.\n' % (epoch - start_epoch + 1, (time.time() - t0) / 3600))
  334. dist.destroy_process_group() if rank not in [-1, 0] else None
  335. torch.cuda.empty_cache()
  336. return results
  337. if __name__ == '__main__':
  338. parser = argparse.ArgumentParser()
  339. parser.add_argument('--weights', type=str, default='yolov5s.pt', help='initial weights path')
  340. parser.add_argument('--cfg', type=str, default='', help='model.yaml path')
  341. parser.add_argument('--data', type=str, default='data/coco128.yaml', help='data.yaml path')
  342. parser.add_argument('--hyp', type=str, default='data/hyp.scratch.yaml', help='hyperparameters path')
  343. parser.add_argument('--epochs', type=int, default=300)
  344. parser.add_argument('--batch-size', type=int, default=16, help='total batch size for all GPUs')
  345. parser.add_argument('--img-size', nargs='+', type=int, default=[640, 640], help='[train, test] image sizes')
  346. parser.add_argument('--rect', action='store_true', help='rectangular training')
  347. parser.add_argument('--resume', nargs='?', const=True, default=False, help='resume most recent training')
  348. parser.add_argument('--nosave', action='store_true', help='only save final checkpoint')
  349. parser.add_argument('--notest', action='store_true', help='only test final epoch')
  350. parser.add_argument('--noautoanchor', action='store_true', help='disable autoanchor check')
  351. parser.add_argument('--evolve', action='store_true', help='evolve hyperparameters')
  352. parser.add_argument('--bucket', type=str, default='', help='gsutil bucket')
  353. parser.add_argument('--cache-images', action='store_true', help='cache images for faster training')
  354. parser.add_argument('--image-weights', action='store_true', help='use weighted image selection for training')
  355. parser.add_argument('--name', default='', help='renames experiment folder exp{N} to exp{N}_{name} if supplied')
  356. parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
  357. parser.add_argument('--multi-scale', action='store_true', help='vary img-size +/- 50%%')
  358. parser.add_argument('--single-cls', action='store_true', help='train as single-class dataset')
  359. parser.add_argument('--adam', action='store_true', help='use torch.optim.Adam() optimizer')
  360. parser.add_argument('--sync-bn', action='store_true', help='use SyncBatchNorm, only available in DDP mode')
  361. parser.add_argument('--local_rank', type=int, default=-1, help='DDP parameter, do not modify')
  362. parser.add_argument('--logdir', type=str, default='runs/', help='logging directory')
  363. parser.add_argument('--log-imgs', type=int, default=10, help='number of images for W&B logging, max 100')
  364. parser.add_argument('--workers', type=int, default=8, help='maximum number of dataloader workers')
  365. opt = parser.parse_args()
  366. # Set DDP variables
  367. opt.total_batch_size = opt.batch_size
  368. opt.world_size = int(os.environ['WORLD_SIZE']) if 'WORLD_SIZE' in os.environ else 1
  369. opt.global_rank = int(os.environ['RANK']) if 'RANK' in os.environ else -1
  370. set_logging(opt.global_rank)
  371. if opt.global_rank in [-1, 0]:
  372. check_git_status()
  373. # Resume
  374. if opt.resume: # resume an interrupted run
  375. ckpt = opt.resume if isinstance(opt.resume, str) else get_latest_run() # specified or most recent path
  376. log_dir = Path(ckpt).parent.parent # runs/exp0
  377. assert os.path.isfile(ckpt), 'ERROR: --resume checkpoint does not exist'
  378. with open(log_dir / 'opt.yaml') as f:
  379. opt = argparse.Namespace(**yaml.load(f, Loader=yaml.FullLoader)) # replace
  380. opt.cfg, opt.weights, opt.resume = '', ckpt, True
  381. logger.info('Resuming training from %s' % ckpt)
  382. else:
  383. # opt.hyp = opt.hyp or ('hyp.finetune.yaml' if opt.weights else 'hyp.scratch.yaml')
  384. opt.data, opt.cfg, opt.hyp = check_file(opt.data), check_file(opt.cfg), check_file(opt.hyp) # check files
  385. assert len(opt.cfg) or len(opt.weights), 'either --cfg or --weights must be specified'
  386. opt.img_size.extend([opt.img_size[-1]] * (2 - len(opt.img_size))) # extend to 2 sizes (train, test)
  387. log_dir = increment_dir(Path(opt.logdir) / 'exp', opt.name) # runs/exp1
  388. # DDP mode
  389. device = select_device(opt.device, batch_size=opt.batch_size)
  390. if opt.local_rank != -1:
  391. assert torch.cuda.device_count() > opt.local_rank
  392. torch.cuda.set_device(opt.local_rank)
  393. device = torch.device('cuda', opt.local_rank)
  394. dist.init_process_group(backend='nccl', init_method='env://') # distributed backend
  395. assert opt.batch_size % opt.world_size == 0, '--batch-size must be multiple of CUDA device count'
  396. opt.batch_size = opt.total_batch_size // opt.world_size
  397. # Hyperparameters
  398. with open(opt.hyp) as f:
  399. hyp = yaml.load(f, Loader=yaml.FullLoader) # load hyps
  400. if 'box' not in hyp:
  401. warn('Compatibility: %s missing "box" which was renamed from "giou" in %s' %
  402. (opt.hyp, 'https://github.com/ultralytics/yolov5/pull/1120'))
  403. hyp['box'] = hyp.pop('giou')
  404. # Train
  405. logger.info(opt)
  406. if not opt.evolve:
  407. tb_writer, wandb = None, None # init loggers
  408. if opt.global_rank in [-1, 0]:
  409. # Tensorboard
  410. logger.info(f'Start Tensorboard with "tensorboard --logdir {opt.logdir}", view at http://localhost:6006/')
  411. tb_writer = SummaryWriter(log_dir=log_dir) # runs/exp0
  412. # W&B
  413. try:
  414. import wandb
  415. assert os.environ.get('WANDB_DISABLED') != 'true'
  416. logger.info("Weights & Biases logging enabled, to disable set os.environ['WANDB_DISABLED'] = 'true'")
  417. except (ImportError, AssertionError):
  418. opt.log_imgs = 0
  419. logger.info("Install Weights & Biases for experiment logging via 'pip install wandb' (recommended)")
  420. train(hyp, opt, device, tb_writer, wandb)
  421. # Evolve hyperparameters (optional)
  422. else:
  423. # Hyperparameter evolution metadata (mutation scale 0-1, lower_limit, upper_limit)
  424. meta = {'lr0': (1, 1e-5, 1e-1), # initial learning rate (SGD=1E-2, Adam=1E-3)
  425. 'lrf': (1, 0.01, 1.0), # final OneCycleLR learning rate (lr0 * lrf)
  426. 'momentum': (0.3, 0.6, 0.98), # SGD momentum/Adam beta1
  427. 'weight_decay': (1, 0.0, 0.001), # optimizer weight decay
  428. 'warmup_epochs': (1, 0.0, 5.0), # warmup epochs (fractions ok)
  429. 'warmup_momentum': (1, 0.0, 0.95), # warmup initial momentum
  430. 'warmup_bias_lr': (1, 0.0, 0.2), # warmup initial bias lr
  431. 'box': (1, 0.02, 0.2), # box loss gain
  432. 'cls': (1, 0.2, 4.0), # cls loss gain
  433. 'cls_pw': (1, 0.5, 2.0), # cls BCELoss positive_weight
  434. 'obj': (1, 0.2, 4.0), # obj loss gain (scale with pixels)
  435. 'obj_pw': (1, 0.5, 2.0), # obj BCELoss positive_weight
  436. 'iou_t': (0, 0.1, 0.7), # IoU training threshold
  437. 'anchor_t': (1, 2.0, 8.0), # anchor-multiple threshold
  438. 'anchors': (2, 2.0, 10.0), # anchors per output grid (0 to ignore)
  439. 'fl_gamma': (0, 0.0, 2.0), # focal loss gamma (efficientDet default gamma=1.5)
  440. 'hsv_h': (1, 0.0, 0.1), # image HSV-Hue augmentation (fraction)
  441. 'hsv_s': (1, 0.0, 0.9), # image HSV-Saturation augmentation (fraction)
  442. 'hsv_v': (1, 0.0, 0.9), # image HSV-Value augmentation (fraction)
  443. 'degrees': (1, 0.0, 45.0), # image rotation (+/- deg)
  444. 'translate': (1, 0.0, 0.9), # image translation (+/- fraction)
  445. 'scale': (1, 0.0, 0.9), # image scale (+/- gain)
  446. 'shear': (1, 0.0, 10.0), # image shear (+/- deg)
  447. 'perspective': (0, 0.0, 0.001), # image perspective (+/- fraction), range 0-0.001
  448. 'flipud': (1, 0.0, 1.0), # image flip up-down (probability)
  449. 'fliplr': (0, 0.0, 1.0), # image flip left-right (probability)
  450. 'mosaic': (1, 0.0, 1.0), # image mixup (probability)
  451. 'mixup': (1, 0.0, 1.0)} # image mixup (probability)
  452. assert opt.local_rank == -1, 'DDP mode not implemented for --evolve'
  453. opt.notest, opt.nosave = True, True # only test/save final epoch
  454. # ei = [isinstance(x, (int, float)) for x in hyp.values()] # evolvable indices
  455. yaml_file = Path(opt.logdir) / 'evolve' / 'hyp_evolved.yaml' # save best result here
  456. if opt.bucket:
  457. os.system('gsutil cp gs://%s/evolve.txt .' % opt.bucket) # download evolve.txt if exists
  458. for _ in range(300): # generations to evolve
  459. if os.path.exists('evolve.txt'): # if evolve.txt exists: select best hyps and mutate
  460. # Select parent(s)
  461. parent = 'single' # parent selection method: 'single' or 'weighted'
  462. x = np.loadtxt('evolve.txt', ndmin=2)
  463. n = min(5, len(x)) # number of previous results to consider
  464. x = x[np.argsort(-fitness(x))][:n] # top n mutations
  465. w = fitness(x) - fitness(x).min() # weights
  466. if parent == 'single' or len(x) == 1:
  467. # x = x[random.randint(0, n - 1)] # random selection
  468. x = x[random.choices(range(n), weights=w)[0]] # weighted selection
  469. elif parent == 'weighted':
  470. x = (x * w.reshape(n, 1)).sum(0) / w.sum() # weighted combination
  471. # Mutate
  472. mp, s = 0.8, 0.2 # mutation probability, sigma
  473. npr = np.random
  474. npr.seed(int(time.time()))
  475. g = np.array([x[0] for x in meta.values()]) # gains 0-1
  476. ng = len(meta)
  477. v = np.ones(ng)
  478. while all(v == 1): # mutate until a change occurs (prevent duplicates)
  479. v = (g * (npr.random(ng) < mp) * npr.randn(ng) * npr.random() * s + 1).clip(0.3, 3.0)
  480. for i, k in enumerate(hyp.keys()): # plt.hist(v.ravel(), 300)
  481. hyp[k] = float(x[i + 7] * v[i]) # mutate
  482. # Constrain to limits
  483. for k, v in meta.items():
  484. hyp[k] = max(hyp[k], v[1]) # lower limit
  485. hyp[k] = min(hyp[k], v[2]) # upper limit
  486. hyp[k] = round(hyp[k], 5) # significant digits
  487. # Train mutation
  488. results = train(hyp.copy(), opt, device)
  489. # Write mutation results
  490. print_mutation(hyp.copy(), results, yaml_file, opt.bucket)
  491. # Plot results
  492. plot_evolution(yaml_file)
  493. print(f'Hyperparameter evolution complete. Best results saved as: {yaml_file}\n'
  494. f'Command to train a new model with these hyperparameters: $ python train.py --hyp {yaml_file}')