volo
volo copied to clipboard
sail-sg
VOLO: Vision Outlooker for Visual Recognition
When use the pre-trained model VOLO-D4-448， the error as flow: Traceback (most recent call last): File "F:/volo-main/main1_all_complete.py", line 416, in main() File "F:/volo-main/main1_all_complete.py", line 168, in main train_loss,train_accuracy=train(train_loader,model, loss_f,optimizer,epoch,args) File...
 Greetings! Thanks for all your inspiring and excellent VOLO work!!! In reading this paper, I get trouble in comprehending the formulation (8), which depicts the complexity of Outlooker Attention....
The Equ.5:  In my opinion, this equ calculates the sum of features in the neighborhood corresponding to (i,j). But in the code: https://github.com/sail-sg/volo/blob/1f67923404d85cb8012a61b35d7eff782fe90cef/models/volo.py#L94-L95 `F.fold(x, output_size=(H, W), ...)` implements another...
Hi, Congratulations on your excellent work and many thanks for making the code public. I have trained a model using the base settings and no token labels: CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 ./distributed_train.sh 8...
I download pre trained model with the link in the document. but when I try to use it, it can't report an error: > hhw@hhw-A01:~/workspace/DeepLearning/VOLO/pretrained_models$ tar -xf d1_384_85.2.pth.tar > tar:...
Traceback (most recent call last): File "main.py", line 949, in main() File "main.py", line 652, in main train_metrics = train_one_epoch(epoch, File "main.py", line 784, in train_one_epoch loss = loss_fn(output, target)...
When I use the pre-trained model d5-448, the following error appears: Traceback (most recent call last): File "F:/volo-main/main1_all_complete.py", line 415, in main() File "F:/volo-main/main1_all_complete.py", line 72, in main load_pretrained_weights(model, './path/to/pretrained/weights/d5_448_87.0.pth.tar',...
Hi, thanks for sharing volo, a nice work. I used bash''' export CUDA_VISIBLE_DEVICES=1,4,5,6 python -m torch.distributed.launch --nproc_per_node=4 main.py "path/to/dataset" \ --model volo_dd2 --img-size 224 \ -b 100 --lr 1.0e-3 --drop-path...
AMP not enabled. Training in float32. Using native Torch DistributedDataParallel. Scheduled epochs: 310 /pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:312: operator(): block: [0,0,0], thread: [15,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out...
Dear authors: Thanks for your wonderful work! the result in Semantic Segmentation task looks nice, thus, could the code and config about the Semantic Segmentation be published? Thanks!
