81 Commits (520f5de6f0febf30dd3545793f528c3478c18299)

Author SHA1 Message Date
  Laughing 956511dafd
fix LR bug 4 years ago
  Glenn Jocher bb3c346916 model.yaml nc inherited from dataset.yaml 4 years ago
  Glenn Jocher df224a0d8f EMA bug fix #279 4 years ago
  Glenn Jocher 3bdea3f697 strip_optimizer() bug fix #253 4 years ago
  Glenn Jocher 597ed4ce63
Update train.py 4 years ago
  Glenn Jocher 9705c15da1
Update train.py 4 years ago
  ChristopherSTAN a8d17ce712
Update train.py 4 years ago
  Glenn Jocher 13f69777a6 typo fix 4 years ago
  yxNONG 1aa2b67933
Update train.py 4 years ago
  Glenn Jocher 86784cfdbf --resume bug fix #252 4 years ago
  Glenn Jocher ad4c22cbfe --resume bug fix #187 4 years ago
  Glenn Jocher 3b16c865f0 assert --epochs 4 years ago
  yxNONG cdb9bde181
Unify the check point of single and multi GPU 4 years ago
  Glenn Jocher b203c9b7ff update train.py incompatible model message fix #222 4 years ago
  Glenn Jocher 37e13f8846 update mosaic border 4 years ago
  Glenn Jocher 22fb2b0c25 refactor dataloader 4 years ago
  Glenn Jocher 256a3e89d2 small dataset bug fix #140 4 years ago
  Glenn Jocher b50fdf16af model.names multi-GPU bug fix #94 4 years ago
  Alex Stoken de191655e4 Fix yaml saving (don't sort keys), reorder --opt keys, bug fix hyp dict accessor 4 years ago
  Alex Stoken 2d396bea00 Fix bug in --help from percent sign in help string 4 years ago
  Glenn Jocher b8557f87e3 add stride to datasets.py 4 years ago
  Alex Stoken 611aacf1bf Turn opt into dictionary before sending it to yaml 4 years ago
  Alex Stoken bc4ef4861b Default optimizer SGD 4 years ago
  Alex Stoken 7abf202cad Mode all optimizer settings to 'hyp.yaml', integrate proper momentum with Adam optimizer 4 years ago
  Alex Stoken 7edbf6570e Fix help message for cfg files 4 years ago
  Alex Stoken d64ad0fbf3 Remove --resume functionality and related checks/logic. 4 years ago
  Glenn Jocher 6c1b87a42e update google_utils import 4 years ago
  Glenn Jocher 1f1917ef56 remove fast, add merge 4 years ago
  Alex Stoken e572bb0803 Add plot_results save location to log_dir 4 years ago
  Lornatang 899f1d4bde Fix DDP bug in single process multiple device use cases 4 years ago
  Glenn Jocher cce95e744d backbone as FP16, save default to FP32 4 years ago
  Glenn Jocher d9b64c27c2 save ckpt in FP16 #119 4 years ago
  Glenn Jocher 9fdb0fbacf AutoAnchor bug fix # 117 4 years ago
  Alex Stoken c8152c81a6 Syntax fixes 4 years ago
  Alex Stoken 9b7386f603 Add save_dir arg to test.test, use arg as location for saving batch jpgs 4 years ago
  Alex Stoken 945307beba Add save_dir to plot_lr_scheduler and plot_labels 4 years ago
  Alex Stoken 3b2b330872 Move results.txt from weights/ to log_dir 4 years ago
  Alex Stoken ade023cff2 Fix hyp file read in and dict update. 4 years ago
  Lornatang 2368603484 fix refrence bug 4 years ago
  Glenn Jocher 8db51c7002 tb_writer bug fix 4 years ago
  Alex Stoken 5f2eeba233 remove old print statements 4 years ago
  Glenn Jocher afe1df385b dist.destroy_process_group() bug fix 4 years ago
  Alex Stoken 333f678b37 add update default hyp dict with provided yaml 4 years ago
  Alex Stoken a448c3bcd7 add logic for resuming and getting hyp for resume run 4 years ago
  Alex Stoken 25e51bcec7 add util function to get most recent last.pt file 4 years ago
  Alex Stoken 490f1e7b9c add save_dir arg to plot_lr_scheduler, default to current dir. 4 years ago
  Alex Stoken 4418809cf5 change weights dir (wdir) to be unique to each run, under log_dir 4 years ago
  Alex Stoken d9f446cd81 add save yaml of opt and hyp to tensorboard log_dir in train() 4 years ago
  Alex Stoken a85e6d0fc0 add parser arg for hyp yaml file 4 years ago
  Glenn Jocher 5a50491fa1 check_anchors bug fix 4 years ago