Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
I
illinois-sl
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Deploy
Releases
Model registry
Monitor
Incidents
Analyze
Value stream analytics
Contributor analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
This is an archived project. Repository and other project resources are read-only.
Show more breadcrumbs
CogComp
illinois-sl
Commits
a7d1f7e5
Commit
a7d1f7e5
authored
9 years ago
by
Shyam Upadhyay
Browse files
Options
Downloads
Patches
Plain Diff
save
parent
3125eadf
No related branches found
Branches containing commit
No related tags found
Tags containing commit
No related merge requests found
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
README.md
+19
-10
19 additions, 10 deletions
README.md
with
19 additions
and
10 deletions
README.md
+
19
−
10
View file @
a7d1f7e5
...
...
@@ -115,8 +115,9 @@ The following script evaluates the performance of 'posModel' on
Use the following comment to train a sequential model 'seqModel' on
'data/tutorial/wsj.sub.train' with DEMIParallelDCD solver:
```
> ./scripts/run_sequence.sh trainSequenceModel data/sequence/wsj.sub.train config/DEMIParallelDCD.config seqModel
```
Again, one can use another structured learning approach to train the model
by specifying a different config file. Each line follows the following format
...
...
@@ -126,8 +127,9 @@ by specifying a different config file. Each line follows the following format
to the same structured example.
The following comment tests 'seqModel" on 'data/sequence/wsj.sub.test' data set.
```
> ./scripts/run_sequence.sh testSequenceModel seqModel data/sequence/wsj.sub.test
```
4.
1.2 __ scripts/run_multiclass.sh __
...
...
@@ -135,8 +137,9 @@ Use the following comment to train a multiclass model 'multiModel' on
'data/multiclass/heart_scale.train' data with a cost matrix specified in
'data/multiclass/heart_scale.cost' using DEMIParallelDCD solver:
```
> ./scripts/run_multiclass.sh trainMultiClassModel data/multiclass/heart_scale.train data/multiclass/heart_scale.cost config/DEMIParallelDCD.config multiModel
```
Each line of the input data represents one instance, and it follows the following format:
[
Tag] [feature1_index
]:
[feature1_value]
[feature2_index]:[feature2_value] ...
...
...
@@ -152,26 +155,31 @@ The cost needs to be positive, and it is 0 when [gold_label]=[predicted_label].
To test the performance of 'multiModel' on 'data/multiclass/heart_scale.test', use
```
> ./scripts/run_multiclass.sh testMultiClassModel multiModel data/multiclass/heart_scale.test
```
4.
1.3 __scripts/run_reranking.sh__
Use the following comment to train a re-ranker on 'data/reranking/rank.train' with Strctured Perceptron
```
> ./scripts/run_reranking.sh trainRankingModel data/reranking/rerank.train config/StructuredPerceptron.config rankModel
```
Use the following comment to test the re-ranker model on 'data/reranking/rerank.test'
```
> ./scripts/run_reranking.sh testRankingModel rankModel data/reranking/rerank.test
```
These scripts should be runnable on installation; However, it is possible,
depending on your system configuration, that you will first need to modify
the permissions on the scripts to allow you to execute them:
```
> chmod 744 scripts/*sh
---------------------- 5 Using the Source Files -----------------------
```
Using the Source Files
-----------------------
COMMAND LINE USAGE
The command lines given at the top of this README.
...
...
@@ -188,7 +196,7 @@ JVM command line parameters.
Additional Documentation/Citation
---------------------
---------------------
Additional documentation is available in the JavaDoc located in doc/index.html
...
...
@@ -203,7 +211,8 @@ K.-W. Chang, V. Srikumar, D. Roth.
Multi-core Structural SVM Training.
ECML, 2013.
----------- 7 Contact Information ------------
Contact Information
------------
Please send a message to illinois-ml-nlp-users@cs.uiuc.edu for any
questions about installing or using the Curator.
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment