Closed rentainhe closed 1 year ago
EMAHook
DINO-R50-4scale-12ep
49.4AP
If you set train.model_ema.enabled=True, after each eval_period, detrex will conduct two evalution on normal model and its ema version.
train.model_ema.enabled=True
eval_period
normal model
ema version
To enabled model ema training:
python tools/train_net.py --config-file projects/dino/configs/dino_r50_4scale_12ep.py \ --num-gpus 8 \ --resume \ train.model_ema.enabled=True \
Testing model with ema states when eval-only:
eval-only
python tools/train_net.py --config-file projects/dino/configs/dino_r50_4scale_12ep.py \ --num-gpus 8 \ --resume \ train.model_ema.enabled=True \ train.model_ema.use_ema_weights_for_eval_only=True \
TODO
EMAHook
as d2go.DINO-R50-4scale-12ep
with ModelEMA:49.4AP
.Note
If you set
train.model_ema.enabled=True
, after eacheval_period
, detrex will conduct two evalution onnormal model
and itsema version
.Usage
To enabled model ema training:
Testing model with ema states when
eval-only
: