mllife
mllife
> > recently, sagemaker have updated the default pytorch kernel to py3.10 with cuda 11.8; so now pyannote is not working properly there > > so what should we do...
try: pipeline = pipeline.to(torch.device("cuda"))
**quickfix**- just change the model 'lp://PubLayNet/faster_rcnn_R_50_FPN_3x/config' to something else example 'lp://PubLayNet/mask_rcnn_X_101_32x8d_FPN_3x/config' and run the code; then change back it will redownload the model again
Yes, both implementations are little different, having the same issue. The newer models on GIT TATR are seems to working better.
I am also seeing the same error, while using the img2img notebook from examples.
I also wanted to express my interest in having these pretrain weights publicly shared. As it is really difficult to train using free resources these days. Publicly sharing the pretrain...
@NormXU @chuuhtetnaing @weishu27
@cau-git , yes, I ran ran a page image through the test code; and then the same page through docling pipeline as a pdf I am seeing different results. I...
all of them, i generated the html using crawler for this page "https://www.bankofcanada.ca/press/"; if i read it from drive again i am getting empty result.links
@cau-git , UPDATE: tried the other backend "pypdfium2", the output is correct now; docling_v2 parser had some text encoding issue