Comments (8)
(a) probably failed because of a check in another script (core/multiprocessed_parsing.py). I changed the check in the newest commit to look at the suffix of the swipl command, so it now also works if you change lines 25/30 in parzu.py to a custom swipl location.
from parzu.
Hi, I fixed this.
When I point to the executable compiled at ~/bin/ I got this error
"[FATAL ERROR: Could not find system resources]"
The error actually occurs twice after the message Starting postprocessor
I fixed this by specifying export SWI_HOME_DIR=/home/elav01/lib/swipl-7.1.26/
on the commandline before running parzu
Now parzu runs without an error, but it still gives an empty output. Any idea how to debug this?
from parzu.
I think you can fix the [FATAL ERROR] issue by also linking to swipl-ld (and maybe also swipl-rc) in your ~/bin directory.
assuming you already pulled my last commit into your git repository, you can try several things to isolate the problem somewhat:
- use '--output preprocessed' or '--output raw' to only perform the first few steps in the pipeline
- process the file 'sample_input' with '--input tagged' to only perform the last few steps in the pipeline
- use '-p 1' to run ParZu single-threaded.
let me know what works and what doesn't.
from parzu.
Hi,
I am trying to use ParZu but I get an empty output. I have read the above posts and tried with --output preprocessed and --output raw and it works. However, with --projective gives me an empty output. Actually, what I need to do is to run the wmt2014-scripts with EMS.
Thank you in advance for your help.
from parzu.
Hi Dinosaxon,
which version of swi-prolog are you using? (and if it's 7.2, do you use the latest version of ParZu?)
there might also be an error message that is printed to a log file (in ParZu/tmp/err*.pl) - so please check that and report if there is anything - you might have to set 'deltemp = 0' in ParZu/config.ini.
If you're interested in having the WMT14/5 data parsed: I put parses for WMT14/15 data on http://statmt.org/rsennrich/parsed_wmt/ , so that you don't have to reparse. You still need to apply the enrich_labelset.py and create_moses_format.py scripts. Then, change the EMS to use 'clean-parsed-stem' instead of 'raw-stem' in the CORPUS section, and 'mock-parsed-corpus' instead of 'raw-corpus' in the LM sections, and point to the new files. I'll try to better document this.
from parzu.
Hi,
Thank you for your prompt reply. Unfortunately, I cannot get the ParZu/tmp/err*.pl. I use the 7.2 version of swi-prolog in Ubuntu 12.04 with the latest ParZu.
I will try to run some more tests and let you know.
Regarding the already parsed sets: my aim is to built my own english-into-german engines so I should resolve this issue. On the other hand they will help me to understand how the output is and confirm that everything is okay.
Once again thank you for your time and consideration. I would appreciate if you can have a look into this even if you don't have more information about my installation.
Thanks
UPDATE: I am really sorry but I have compiled ParZu about a week ago, so that's why the bug. I have updated your code and it works now :)! Thanks
from parzu.
Hi,
I seem to be getting the same error I have an old version of swipl (version 6.6.6, unfortunately I could not install 7.2.x on my machine due to some install errors)
I have linked swipl, swipl-ld and swipl-rc.
$>"Das is ein Test" | ./parzu
-bash: Das is ein Test: command not found
Starting tokenizer
Starting POS-tagger
Starting preprocessor
Tokenizer v3
Language: de
* Load model
* Label sequences
* Done
reading transducer from file "/home/arenduc1/source-pkgs/ParZu/external/zmorge-20140521-smor_newlemma.ca"...
finished.
Starting parser
Starting postprocessor
[FATAL ERROR:
Could not find system resources]
[FATAL ERROR:
Could not find system resources]
[FATAL ERROR:
Could not find system resources]
[FATAL ERROR:
Could not find system resources]
[FATAL ERROR:
Could not find system resources]
from parzu.
You'll have to try to fix your swipl installation before running ParZu.
Here's the relevant entry from the swipl homepage: http://www.swi-prolog.org/FAQ/FindResources.html
from parzu.
Related Issues (20)
- Executing "create_statistics.sh" HOT 3
- Using the python module HOT 1
- Cannot parse file in docker HOT 2
- Difference Local and WebVersion results HOT 5
- ImportError: No module named pexpect when starting docker run HOT 4
- Wrong Tagging for APZR HOT 7
- swi-prolog sfst HOT 1
- Support --tokenized or --tokenized_lines in the REST API HOT 3
- For python scripts to be invoked from command line, consider #!/usr/bin/env python
- Problem running the create_statistics.sh script HOT 3
- Error POS Tatting PPOSAT
- Cannot use the module HOT 1
- Online parser returns random sentences HOT 6
- Timeout values cannot be configured HOT 3
- Some texts take very long to parse HOT 3
- Error 502 with the demo of the ParZu HOT 4
- pexpect error on init HOT 2
- broken link in readme HOT 1
- Unexpected keyword argument 'text' for Python3.1 HOT 1
- Creating the visual dependency graph from conll data HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from parzu.