abdksyed

+1 Same here, when using `Katex` the math is nicely centered, but no italics and small font, when using null, the italics and fonts are good, but the equations are...

Any updates on fine tuning available models on custom dataset??

Thanks, Will look into top-k predictions, will run the inference with that and will get back here with results.

Any update guys, using LayoutXLM separately just for linking would not make sense, and for `Semantic Entity Recognition`, LayoutLMv2 looks better than XLM when looking at numbers on FUNSD dataset....

Nope, the last time I checked, it didn't work. But I don't know after almost 18 months if there is any change or not.

@SrinivasPrashanthInf You will have to use the Faster RCNN model and config using the [detectron2](https://github.com/facebookresearch/detectron2)

@glenn-jocher Is it required to convert it to intermediate Bounding Box format, even if we want to train a YOLO model for segmentation?

Any updates on this @RangiLyu ?

No, the Validation Loss is not supported yet. The only way to check the overfitting is by looking at mAP scores over trained and validation data. There is a way...

@tomhruby1 @willcray ```python @HOOKS.register_module() class FindIoU(Hook): def __init__(self, name): os.makedirs("bestepochs", exist_ok=True) # Some Necessary Variables for me self.bestIoU = 0 self.bestepoch = None self.name = name self.metric = BinaryJaccardIndex() #...