Andy Friedman comments

Results 15 comments of


                                            Andy Friedman

how to run the paraphraser? meet with some problem

does anyone have a mirror for the zip or has it been re-upped somewhere? it's over download cap rn!

Closing tabs in Sequel Pro causes crash "Sequel Pro quit unexpectedly", been like this forever. It's super annoying, so you'll have to leave every single tab open, sometimes it's dozens of them

came here to say this, it's super annoying

Allow translation of invitation emails

hey -- I'm working on this and having a little trouble figuring out an Jinja error I did this ... 1. Renamed "invitation_mail.en.j2" to "invitation_mail.j2" 2. Added `{% trans %}`...

Allow translation of invitation emails

found a workaround! pull request is up.

AttributeError: module 'deepdoctection.extern' has no attribute 'D2FrcnnDetector'

had the same problem, got it working with a fresh environment here's the requirements.txt i used... ```torch==2.2.1 torchvision python-doctr deepdoctection pdfplumber autopep8 isort ipykernel pango```

For text extraction, add fractional versions of `x/y_tolerance` arguments

is anyone working on this? I'd like to take a crack at it if not!

For text extraction, add fractional versions of `x/y_tolerance` arguments

great -- do you have a pdf with extra-tight letters I can use for testing? (the pdf in the original issue worked fine with `x_toleraance=1` sooo....)

For text extraction, add fractional versions of `x/y_tolerance` arguments

sorry im confused -- what do we want it to output?

For text extraction, add fractional versions of `x/y_tolerance` arguments

is there a less dumb way to get text size than `int(text['bottom'] - text['top'])`?

For text extraction, add fractional versions of `x/y_tolerance` arguments

Honestly I'm lazy and couldn't find an easy way to extract `char` objects from text, so now it works with anything with a `top` and `bottom` param. The `int` isn't...