Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

./Inference/README.md use the same checkpoint for R50 and R50IBN ? #39

Open
taqta opened this issue Jun 7, 2023 · 0 comments
Open

./Inference/README.md use the same checkpoint for R50 and R50IBN ? #39

taqta opened this issue Jun 7, 2023 · 0 comments

Comments

@taqta
Copy link

taqta commented Jun 7, 2023

it seems that creating_embeddings.py use R50 model, and use "logs/market1501/256_resnet50/train_ctl_model/version_0/checkpoints/epoch=119.ckpt" as ckpt in README.md

python inference/create_embeddings.py \
--config_file="configs/256_resnet50.yml" \
GPU_IDS [0] \
DATASETS.ROOT_DIR '/data/my-image-dataset/' \
TEST.IMS_PER_BATCH 128 \
OUTPUT_DIR 'output-dir' \
TEST.ONLY_TEST True \
MODEL.PRETRAIN_PATH "logs/market1501/256_resnet50/train_ctl_model/version_0/checkpoints/epoch=119.ckpt"

and get_similar.py use R50_IBN as model, and use "logs/market1501/256_resnet50/train_ctl_model/version_0/checkpoints/epoch=119.ckpt" as ckpt in README.md

python inference/get_similar.py \
--config_file="configs/320_resnet50_ibn_a.yml" \
--gallery_data='output-dir' \
--normalize_features \
--topk=100 \
GPU_IDS [0] \
DATASETS.ROOT_DIR '/data/query-images/'  \
TEST.IMS_PER_BATCH 128 \
OUTPUT_DIR 'output-dir' \
TEST.ONLY_TEST True \
MODEL.PRETRAIN_PATH "logs/market1501/256_resnet50/train_ctl_model/version_0/checkpoints/epoch=119.ckpt" \
SOLVER.DISTANCE_FUNC 'cosine'

I want to ask that why the same checkpoint can be used in R50 and R50_IBN, and does it mean that R50_IBN use the default mean and var if I run the prompt as README.md.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant