The kelpie from androssi

How do you freeze all the other parameters except for the kelpie entity embedding?

Hello, Andrea~ thank you for your excellent work! Currently, I am learning your paper and codes.
I am wondering how or where you freeze all the other parameters during post-training. I checked the codes, especially in the optimizer class . It seems like all embeddings are updated during post-training?

Looking forward to your reply!

Minor availability issues

Hi,

Thank you for the very detailed and self-contained scripts for availability/reproducibility. There are a couple of issues I have run into:

There are absolute paths in reproducibility_generate_pdf.sh which depend on the location of the cloned repo
After fixing the absolute paths, there is an issue with generate_pdf_report.py while running reproducibility_generate_pdf.sh the backtrace is:

Traceback (most recent call last):
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/lib/utils.py", line 523, in open_for_read
    return open_for_read_by_name(name,mode)
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/lib/utils.py", line 463, in open_for_read_by_name
    return open(name,mode)
FileNotFoundError: [Errno 2] No such file or directory: '/home/nicholso/tmp/Kelpie/reproducibility_images/extraction_times_with_and_without_prefilter_plot.png'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/lib/utils.py", line 530, in open_for_read
    return BytesIO((datareader if name[:5].lower()=='data:' else rlUrlRead)(name))
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/lib/utils.py", line 476, in rlUrlRead
    return urlopen(name).read()
  File "/usr/lib/python3.8/urllib/request.py", line 222, in urlopen
    return opener.open(url, data, timeout)
  File "/usr/lib/python3.8/urllib/request.py", line 509, in open
    req = Request(fullurl, data)
  File "/usr/lib/python3.8/urllib/request.py", line 328, in __init__
    self.full_url = url
  File "/usr/lib/python3.8/urllib/request.py", line 354, in full_url
    self._parse()
  File "/usr/lib/python3.8/urllib/request.py", line 383, in _parse
    raise ValueError("unknown url type: %r" % self.full_url)
ValueError: unknown url type: '/home/nicholso/tmp/Kelpie/reproducibility_images/extraction_times_with_and_without_prefilter_plot.png'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/lib/utils.py", line 643, in __init__
    fp = open_for_read(fileName,'b')
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/lib/utils.py", line 532, in open_for_read
    raise IOError('Cannot open resource "%s"' % name)
OSError: Cannot open resource "/home/nicholso/tmp/Kelpie/reproducibility_images/extraction_times_with_and_without_prefilter_plot.png"

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "generate_pdf_report.py", line 191, in <module>
    generate_pdf()
  File "generate_pdf_report.py", line 187, in generate_pdf
    doc.build(flowables)
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/platypus/doctemplate.py", line 1317, in build
    BaseDocTemplate.build(self,flowables, canvasmaker=canvasmaker)
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/platypus/doctemplate.py", line 1082, in build
    self.handle_flowable(flowables)
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/platypus/doctemplate.py", line 931, in handle_flowable
    if frame.add(f, canv, trySplit=self.allowSplitting):
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/platypus/frames.py", line 169, in _add
    w, h = flowable.wrap(aW, h)
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/platypus/flowables.py", line 511, in wrap
    return self.drawWidth, self.drawHeight
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/platypus/flowables.py", line 505, in __getattr__
    self._setup_inner()
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/platypus/flowables.py", line 462, in _setup_inner
    img = self._img
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/platypus/flowables.py", line 499, in __getattr__
    self._img = ImageReader(self._file)
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/lib/utils.py", line 655, in __init__
    annotateException('\nfileName=%r identity=%s'%(fileName,self.identity()))
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/lib/utils.py", line 1176, in annotateException
    rl_reraise(t,t(sep.join((_ for _ in (msg,str(v),postMsg) if _))),b)
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/lib/utils.py", line 138, in rl_reraise
    raise v.with_traceback(b)
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/lib/utils.py", line 643, in __init__
    fp = open_for_read(fileName,'b')
  File "/home/nicholso/.local/lib/python3.8/site-packages/reportlab/lib/utils.py", line 532, in open_for_read
    raise IOError('Cannot open resource "%s"' % name)
OSError:
fileName='/home/nicholso/tmp/Kelpie/reproducibility_images/extraction_times_with_and_without_prefilter_plot.png' identity=[ImageReader@0x7f3d4e4a66a0 filename='/home/nicholso/tmp/Kelpie/reproducibility_images/extraction_times_with_and_without_prefilter_plot.png'] Cannot open resource "/home/nicholso/tmp/Kelpie/reproducibility_images/extraction_times_with_and_without_prefilter_plot.png"

based on the contents of reproducibility_generate_pdf.sh, I think there is a missing call to the script which generates extraction_times_with_and_without_prefilter_plot.png

At the end of the running paper experiments section of the readme there is a minor typo in which script to run. I believe So after running the script, is is sufficient to re-run the PDF generation script reproducibility_environment.sh to obtain an up-to-date PDF report. should be So after running the script, is sufficient to re-run the PDF generation script sh reproducibility_generate_pdf.sh to obtain an up-to-date PDF report.

I am running this on Ubuntu 20.04 with python 3.8.10 in a clean venv, only installing packages through reproducibility_environment.sh. If you need any additional information to reproduce point 2 let me know.

Would you be able to take a look at these and update the scripts if necessary?

Thank you,
Hamish

Images thread

This is not a real issue, just a way to post images so that they can be easily referenced in the README.md

External KGE Models (from pykeen)

Hello, Andrea,

Thank you for sharing this amazing work !
I am wondering if it's possible to use externel .pkl/pth from another framework like pykeen. In fact, I've already trained some models from pykeen on my own data and i'm looking for an XAI tool, and I find your work.

Looking forward to your reply!

Gaoussou

androssi / kelpie Goto Github PK

kelpie's People

Contributors

Stargazers

Watchers

Forkers

kelpie's Issues

How do you freeze all the other parameters except for the kelpie entity embedding?

Minor availability issues

Images thread

External KGE Models (from pykeen)

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent