Why torch.compile has very small acceleration for Donut model?
#10 opened 4 days ago
by
gorodnitskiy
Will this modle suitable for invoce processing ?
1
#9 opened 4 months ago
by
total008
Discrepancies between DONUT / BART Tokenizer and missing characters
1
#8 opened 7 months ago
by
DieseKartoffel
Adding `safetensors` variant of this model
#7 opened 9 months ago
by
SFconvertbot
Architecture of donut
#6 opened 10 months ago
by
shubham05
Change image_mean and image_std to ImageNet to match original codebase
#5 opened over 1 year ago
by
morgan
Failed to download Donut Processor
5
#3 opened over 1 year ago
by
paturi1710
Minimum GPU requirement to fine-tune donut model
#2 opened almost 2 years ago
by
LeonardoVaz