Commit History
Update train.py (#902)
281d78c
unverified
Remove autoanchor and class checks on resumed training (#889)
6f3db5e
unverified
opt.image_weights bug fix (#885)
f06e2d5
opt.img_weights bug fix (#885)
69ff781
Update hyperparameters to add lrf, anchors
08e97a2
Update train.py forward simplification
a21bd06
torch.from_tensor() bug fix
09402a2
remove ema.ema hasattr(ema, 'module') check
83dc540
--resume to same runs/exp directory (#765)
4447f4b
unverified
Fix ema attribute error in DDP mode (#775)
fb4fc8c
unverified
single command --resume (#756)
ebafd1e
unverified
v3.0 Release (#725)
916d4aa
unverified
Fix Logging (#719)
0892c44
unverified
max workers for dataloader (#722)
a925f28
unverified
Marc
commited on
Fix redundant outputs via Logging in DDP training (#500)
4949401
unverified
Model freeze capability (#679)
e71fd0e
update train.py remove save_json final_epoch
8e5c665
Dataset autodownload feature addition (#685)
41523e2
unverified
Fix missing model.stride in DP and DDP mode (#683)
3d8ed0a
unverified
Single-source training update (#680)
a0ac5ad
Single-source training (#680)
3c6e2f7
unverified
Fix unrecognized local rank argument (#676)
d7cfbc4
unverified
train.py --logdir argparser addition (#660)
9368453
unverified
Add Multi-Node support for DDP Training (#504)
886b984
unverified
Fix missing imports (#627)
728efa6
unverified
import random bug fix (#614)
eb99dff
Explicit Imports (#498)
d5b6416
unverified
Jirka Borovec
commited on
hyperparameter evolution update (#566)
f1096b2
hyperparameter evolution bug fix (#566)
c1a2a7a
hyperparameter evolution bug fix (#566)
8074745
hyperparameter evolution bug fix (#566)
e32abb5
hyperparameter evolution bug fix (#566)
8056fe2
hyperparameter expansion to flips, perspective, mixup
127cbeb
replace random_affine() with random_perspective()
bcd452c
PyTorch 1.6.0 update with native AMP (#573)
c020875
unverified
LR --resume repeat bug fix (#565)
4e2b9ec
unverified
Laughing
commited on