Kirill Kozlov
Kirill Kozlov
> Does it have a warning in the beginning? No
> What's the training code/command you're using? yolo segment train data=... model=best.pt batch=4 project=grid_seg_exp/ overlap_mask=False augment=True epochs=30 patience=15 mosaic=False imgsz=1920 device=0 plots=True exist_ok=True workers=20 amp=False scale=0 pretrained=True
> There's probably an issue with your dataset labels. Does it have boxes with no masks? No, in this case I'd have a problem with a simple training without flag....
> Is the error immediately after the training starts or during some random iteration or during validation? during training on first epoch
yeah, I will try with last version, thx
Well, I guess the error was occurred by my addition to augmentations, without it it works, however, overlap_mask=False doesn't resolve problem, I don't understand why for a particular box a...
> Running the ONNX model on a sample image results in a latency of around 60ms. When converting the ONNX model to TensorRT format, I get a latency of 70–80ms...
> [@KozlovKY](https://github.com/KozlovKY) Thank you for your response. > > I'm still encountering issues achieving the proposed 2-3ms latency for the small model using TensorRT. My TensorRT inference time remains around...