Please provide the following information when requesting support.
• Hardware: NVIDIA GeForce RTX 4090
• Network Type: Detectnet_v2
• TAO Version: 5.2.0
• Training spec file: detectnet_v2_train_resnet34_kitti.txt (6.2 KB)
The sample training file I use is the notebooks/tao_launcher_starter_kit/detectnet_v2/ Detectnet_v2.ipynb provided in the TAO ToolKit Getting Started. I had no problem training with resnet18, I just modified to resnet34 and [ERROR] tensorflow occurred.
!tao model detectnet_v2 train -e $SPECS_DIR/detectnet_v2_train_resnet34_kitti.txt \
-r $USER_EXPERIMENT_DIR/experiment_dir_unpruned \
-n resnet34_detector \
--gpus $NUM_GPUS
The following is the output after the above command is run.
log.txt (623.4 KB)
2 posts - 2 participants
