-
using well-known DL-models like ImageNet, GoogleNet and AlexNet we can detect an object like a small cat or a face in a bigger image, and mark the detected object by a box.
With tiny_dnn we want to us…
-
Hi. I downloaded the imagenet1k file from huggingface, which has the train_images_0.tar.gz file. However, whenever I extract it (by running tar -xf train_images_0.tar.gz), I only get images, and not s…
-
I have test the model on ImageNet-1k val set with zero-shot setting and the labels are translated to Chinese. However the top1 accuracy is only around 25%. As a comparison, the digit on CLIP is 65%.
…
-
Hello dude, I am a beginner attempting to reproduce your article. However, I have found that the accuracy and training curves on Wandb are very poor. Could you please advise if this issue is due to my…
-
-
## TL; DR
- ViT feature representations are *less hierarchical*.
- Early tr blocks learn both local and global dependencies provided with large enough dataset.
- Skip connections play much more i…
-
Hi, there! I'm very interested in this project, but when I reproduced this work, I found that the script you provided in meta.ipynb may be incomplete as the variable 'unlab_x' is not predefined. Could…
-
Hi, I'm training faster-rcnn on 4 gpus with coco dataset converted to LMDB.
I used num_worker=4 for the dataloader and I found that the memory occupation is almost 60Gb.
I suspect that the whole dat…
-
sorry to leave a question like this here, but i saw your issue at https://github.com/blue-blue272/fewshot-CAN/issues/4. May i ask did you solve the tieredimagenet dataset and ran the code? If you did,…
-
I re-trained the model as per the README running:
`torchrun --nproc_per_node=8 --master_port=22447 --max_restarts=0 train.py \
--model_name microsoft/Phi-3.5-vision-instruct --bf16 --pooling last…