profile
viewpoint
Becky Sharp BeckySharp Tucson, AZ

BeckySharp/OdinsonWebapp 3

Odinson development webapp

WorldModelers/Ontologies 3

Ontologies for the World Modelers system

enoriega/cs665 1

Repository for question answering for CSC 665

BeckySharp/data 0

Databases & Data Science Playground

BeckySharp/deep_qa 0

Using deep learning to answer Aristo's science questions

BeckySharp/eidos 0

Protests Reading system

pull request commentclulab/eidos

Move SentenceClassifier to EidosComponents

LGTM, but maybe @BeckySharp should say it.

fine with me too!

kwalcock

comment created time in 10 hours

MemberEvent
MemberEvent

issue openedlum-ai/odinson

debug mode for rule learning

As a rule-learning user I want to be able to enable/disable the path-keeping in the match, so that I can use it for rule-learning when I need it, but can turn it off for efficiency when I don't.

I think the path-remembering code is in the path branch, so once this is in place we need to merge that into master. This will allow rule-learning from the master branch.

created time in 15 hours

push eventclulab/asist-speech

bsharpataz

commit sha 0d55ab0208d7c062e16254ec3d01ca7b23c8e91c

learning

view details

push time in 3 days

push eventclulab/asist-speech

bsharpataz

commit sha 19c3482218eb9548484b21058418a60219ac22e0

lots of small changes

view details

push time in 4 days

PR opened ml4ai/automates

Reviewers
Eqn cleanup

@cl4yton

This is documentation on the scripts in the equation extraction/decoding pipeline. while there are other scripts, basically if it's not mentioned it's not relevant to any future work. I moved things around. Though I did it with PyCharm (which was refactoring as I went), there's a chance that some imports/paths are now not working, but whoever is working on this in the future can fix that if it's an issue. These are somewhat special purpose, and the plan changed as we went. For some of these it really may be easier for whoever to start over, and that's fine. Hope it helps!

+87 -30

0 comment

29 changed files

pr created time in 4 days

push eventml4ai/automates

bsharpataz

commit sha 68a294bbc858200ceb404721377eb3baf1b824d9

oops

view details

push time in 4 days

push eventml4ai/automates

bsharpataz

commit sha 02f9ac44b769fd3935de622000b63da1d5e7485f

cleanup and reorg

view details

push time in 4 days

issue openedlum-ai/odinson

Add scala docs

We need scala docs to clearly document the functionality of each Odinson class. These should be linked to the current docs.

created time in 4 days

issue openedlum-ai/odinson

Remove handling of single quoted strings

We currently support single and double quotes around strings, but we are using java's approach to escape characters in the string, which is designed for double quote only. Rather than jump through flaming hoops, we should just no longer support single quotes.

TODO:

  • [ ] remove handling in the QueryParser
  • [ ] edit the documentation String page

created time in 5 days

issue closedlum-ai/odinson

Test single quote

Testing

While we allow either single quotes or double quotes to wrap strings with special stuff going on: e.g., "3:10" to Yuma vs. '3:10' to Yuma

Behind the scenes, we're using java escaping rules, which doesn't allow single quotes. We need to see if the single quote is working, specifically with escaped characters inside a single quote. e.g., 'abc"def' or something with a unicode escaped char.

@marcovzla please add examples/clarify the details of this issue. thanks!

closed time in 5 days

BeckySharp

create barnchml4ai/automates

branch : eqn_cleanup

created branch time in 5 days

create barnchclulab/asist-speech

branch : clara-bsharp

created branch time in 5 days

push eventclulab/asist-speech

Becky Sharp

commit sha 133147c713eca14de6f032e7f00b28dfe5e3f0ef

comments

view details

push time in 6 days

PR opened clulab/asist-speech

Reviewers
added comment

not done yet...

+2 -1

0 comment

1 changed file

pr created time in 6 days

create barnchclulab/asist-speech

branch : BeckySharp-comments

created branch time in 6 days

issue commentlum-ai/odinson

Add info to docs

ps -- docs are here: http://gh.lum.ai/odinson/

BeckySharp

comment created time in 6 days

issue openedlum-ai/odinson

Add info to docs

End-user documentation

We need to add (not necessarily in this order, though I did try to put them in a semblance of order):

  • [ ] rest-api info
  • [ ] docker info
  • [ ] event query info
  • [ ] state (once it's ready)
  • [ ] the yaml format / working with rule files
  • [ ] as part of yaml format -> variables
  • [ ] walk-through example (ref external webapp)
  • [ ] walk-through example in code
  • [ ] structure of a Mention
  • [ ] accessing things like named captures, etc.
  • [ ] structure of an OdinsonDocument

We also need to:

  • [ ] re-write the main landing README page to be simpler and point people to docs

others?

Proofreading/editing of current content SUPER appreciated, but @MihaiSurdeanu only gets half a vote about the colors bc I don't trust his color taste (since he doesn't like my eidos spinner) :D mostly kidding... but I really do like the colors I picked...

created time in 6 days

delete branch lum-ai/odinson

delete branch : docs_edits

delete time in 6 days

push eventlum-ai/odinson

Becky Sharp

commit sha 04a084b37a313b527648fea15eec649619afca82

Docs edits (#106) * progress on docs * one more todo

view details

push time in 6 days

PR merged lum-ai/odinson

Docs edits

addresses comments from mihai, adds extra content, modified the colors (super fun!), etc.

+120 -36

0 comment

12 changed files

BeckySharp

pr closed time in 6 days

PR opened lum-ai/odinson

Docs edits

addresses comments from mihai, adds extra content, modified the colors (super fun!), etc.

+120 -36

0 comment

12 changed files

pr created time in 6 days

push eventlum-ai/odinson

bsharpataz

commit sha 1e00a31f251316531716bdcfd6bdaaf90d6c74c7

one more todo

view details

push time in 6 days

create barnchlum-ai/odinson

branch : docs_edits

created branch time in 6 days

delete branch marcovzla/discobert

delete branch : smallFix

delete time in 6 days

push eventmarcovzla/discobert

maxaalexeeva

commit sha fa88b83df9d5d47c09dbe6a83718df11444bd954

changed dimension---dim=0 is always 1 for the scores tensor tensor([[-0.0055, 0.1544]])

view details

maxaalexeeva

commit sha bcb14bdb0255a90d4a437cc10600cde2d4563029

maskig fix

view details

Becky Sharp

commit sha fa2cae179c9bdf8e61f557ab4324a6646c80a165

Merge pull request #4 from marcovzla/smallFix some dimension quirks

view details

push time in 6 days

PR merged marcovzla/discobert

Reviewers
some dimension quirks

I changed a dimension in best_legal_action because I think the size of dim=0 is always 1 for the scores tensor tensor([[-0.0055, 0.1544]])

This did not influence the results from classifier with two actions: if there was only one legal action, that is the action that got returned: https://github.com/marcovzla/discobert/blob/2e6f1fe0d0cd508bfa97de480a2fe40e8c27f67d/model.py#L76

if there were two legal actions, both of them got masked correctly (for some reason---more on that below): https://github.com/marcovzla/discobert/blob/2e6f1fe0d0cd508bfa97de480a2fe40e8c27f67d/model.py#L86

This is only an issue if there are more than 2 actions.

Also, this masking

https://github.com/marcovzla/discobert/blob/2e6f1fe0d0cd508bfa97de480a2fe40e8c27f67d/model.py#L86

does not seem to work correctly if the number of items in the list to mask is not equal to the number of items to mask? Here are some outputs from when there are three actions:

masked correctly:

scores:  tensor([[ 0.3933, -0.3062,  0.1215]], device='cuda:0',
       grad_fn=<UnsqueezeBackward0>)
action ids: [0, 1, 2]
mask:  tensor([[0., 0., 0.]], device='cuda:0')

masked correctly:

scores:  tensor([[ 0.2237, -0.2750,  0.0463]], device='cuda:0',
       grad_fn=<UnsqueezeBackward0>)
action ids: [0, 1, 2]
mask:  tensor([[0., 0., 0.]], device='cuda:0')

masked wrong:

scores:  tensor([[ 0.2023, -0.2509,  0.3609]], device='cuda:0',
       grad_fn=<UnsqueezeBackward0>)
action ids: [1, 2]
mask:  tensor([[-inf, -inf, -inf]], device='cuda:0')

I tried to replicate this in the shell, but I can't. Trying to mask this way, I get the index out of bounds error, but the mask is still updated. There's also a good chance I am just missing something about how tensors work, but I can't track it down. Could have to do with the fact that action scores have been unsqueezed? I'm quite confused.

In one of my other branches, where this is crucial, I opted for now for a for-loop. If the loop looks right to you, I could add it with this PR. Here's the loop:

for i in action_ids:
    mask[0][i] = 0
+2 -2

4 comments

1 changed file

maxaalexeeva

pr closed time in 6 days

pull request commentmarcovzla/discobert

some dimension quirks

cool, i'll take the liberty of merging then... ("don't mind if I do!") :)

maxaalexeeva

comment created time in 6 days

delete branch lum-ai/odinson

delete branch : offline-docs

delete time in 6 days

push eventlum-ai/odinson

George C. G. Barbosa

commit sha 69d4409350d0e79d84f564394865e04377aaf860

Offline docs (#105) * Ignore docs temp files * Add Gemfile to bundle and generate offline docs * Add description on how to deploy docs offline * Add theme for offline usage * Add _site back. Idk if it was ok to ignore it

view details

push time in 6 days

PR merged lum-ai/odinson

Offline docs
  • Add a Gemfile for offline deployment
  • Add a doc page describing how to deploy docs offlihne
  • Ignore docs temp files
+319 -0

1 comment

5 changed files

gcgbarbosa

pr closed time in 6 days

issue openedlum-ai/odinson

Testing examples in the Docs

Testing

Now that the docs are live: http://gh.lum.ai/odinson/ We should make a unit test from each of the examples provided.
There are currently several examples, likely some more will be added, but there are enough to get started.

Thanks!

created time in 7 days

issue openedlum-ai/odinson

Test single quote

Testing

While we allow either single quotes or double quotes to wrap strings with special stuff going on: e.g., "3:10" to Yuma vs. '3:10' to Yuma

Behind the scenes, we're using java escaping rules, which doesn't allow single quotes. We need to see if the single quote is working, specifically with escaped characters inside a single quote. e.g., 'abc"def' or something with a unicode escaped char.

@marcovzla please add examples/clarify the details of this issue. thanks!

created time in 7 days

issue openedlum-ai/odinson

Testing the ! operator

Testing

Test the ! operator for token constraints.

E.g., [tag=/N.*/ & !lemma=puppy] (I may have spaces wrong)

created time in 7 days

issue closedlum-ai/odinson

Testing the ! operator

Documentation type: [End-user documentation, wiki, code comments]

Improvements requested:

Link (if available):

closed time in 7 days

BeckySharp

issue openedlum-ai/odinson

Testing the ! operator

Documentation type: [End-user documentation, wiki, code comments]

Improvements requested:

Link (if available):

created time in 7 days

push eventlum-ai/odinson

Becky Sharp

commit sha e00215c296f83f99ff98768ae7def79c89c89ded

jekyll (#98)

view details

Becky Sharp

commit sha ee18895b671885f926bda3f439a09bfdd5b6e6c4

Patch (#99) * jekyll * endraw

view details

Becky Sharp

commit sha fad23e72d9c5ec5ee2b9f3a8bbb140a4dca38249

Merge branch 'master' into patch

view details

push time in 7 days

PR opened lum-ai/odinson

Reviewers
Patch

:) some tweaks

+28 -84

0 comment

7 changed files

pr created time in 7 days

push eventlum-ai/odinson

bsharpataz

commit sha 21118d6e944b87e12206cf9b4c68b1c42713312f

trying link color change

view details

push time in 7 days

create barnchlum-ai/odinson

branch : patch

created branch time in 7 days

push eventlum-ai/odinson

Becky Sharp

commit sha e00215c296f83f99ff98768ae7def79c89c89ded

jekyll (#98)

view details

Becky Sharp

commit sha 0780a40b4372b096cc11ff0b560c7b5921cbc9cf

Merge branch 'master' into patch

view details

push time in 7 days

PR opened lum-ai/odinson

Reviewers
Patch
+4 -1

0 comment

1 changed file

pr created time in 7 days

create barnchlum-ai/odinson

branch : patch

created branch time in 7 days

PR opened lum-ai/odinson

Reviewers
jekyll

patch for ghpages string replacement

+4 -1

0 comment

1 changed file

pr created time in 7 days

create barnchlum-ai/odinson

branch : patch

created branch time in 7 days

push eventlum-ai/odinson

bsharpataz

commit sha 5064c7586e4d9e1b403935056d873401b16436bb

added docs to changes

view details

push time in 7 days

PR opened lum-ai/odinson

Doc draft

documentation draft

+370 -1

0 comment

15 changed files

pr created time in 7 days

push eventlum-ai/odinson

bsharpataz

commit sha d765503c9344b1843cf79b558acc0cf1a9d538af

edits

view details

push time in 7 days

issue openedlum-ai/odinson

Annotation agnosticism

As a user with a custom index, I would like to include arbitrary token-level tags.

I think we need to adjust the code that interacts with this portion of the config: https://github.com/lum-ai/odinson/blob/0646f4a8ab460a6bc9b4a513457fcc42fbaf46a2/core/src/main/resources/reference.conf#L33-L103

Specifically the QueryCompiler... (e.g., https://github.com/lum-ai/odinson/blob/0646f4a8ab460a6bc9b4a513457fcc42fbaf46a2/core/src/main/scala/ai/lum/odinson/compiler/QueryCompiler.scala#L564 ++)

Also -- while we're making changes, let's rename "dependenciesField" to "graphField" since we're trying to be all annotation agnostic and stuff. :)

created time in 7 days

push eventlum-ai/odinson

marcovzla

commit sha a7d838d9932820106ed3bd7db2023209172714d1

size of roots array (#74) * size of int * changes

view details

Gus Hahn-Powell

commit sha bd444f8700b9cbe4b86fa6f50fe0e18d40328605

Containerized Odinson (#76) * Containerized Odinson - images for `extra` (`lumai/odinson-extras`) and REST API (`lumai/odinson-rest-api`) * Updated CHANGES.md * Version bump and adjustments to dispel warnings

view details

Gus Hahn-Powell

commit sha bd48c0ff4f63a392075fe76648acab09f0134b00

Improvements to the REST API (#77) * Resolves #70. * API endpoints for retrieving metadata, OdinsonDocument, etc. * Updates to OpenAPI specification (endpoints and data models) * Sentence -> OdinsonSentence

view details

George C. G. Barbosa

commit sha e2451e989ab02805ffd44506cc55dc03f11431e5

**testing codcov** (#81) * Add code-coverage plugin * Add git action to push results to codecov * Fix indentation * Pretty * Delete scala.yml * Trying again * Update scala.yml * Change CHANGES.md * Change sbt version back to 1.2.6 * Update README.md

view details

Keith Alcock

commit sha d58a6efd0a2dbe3644a0fcdecd03bd75cd32000c

Start State trait (#82) * Turn State into a trait Convert HikariDataSource into first implementation * Push to origin in case anyone looks * Rename to general SqlState, add FileState placeholder * Implement simplest possible MemoryState * Implement addMentions * Make second index * Add comment about second table * Tweek configuration files Change reference.conf back to sql. Add some missing entries into application.conf so that extras work. * Add counter to get tests to pass in sbt Tests will not pass if run twice in one sbt session * Update CHANGES.md

view details

George C. G. Barbosa

commit sha 12f3d82acb261c2c7023a51d324b7a9a218e558a

Organizing core tests (#80) * Ignore georges files * Add part of the description of the rules file * Added the rest of the description of the rules * Update README.md * Move serialization and traversals stuff * Start list of fixtures * Add more observations from scalaTest doc * Ignore ctags * Add one more category of tests * Move basic tests * Move events tests * Moved patterns tests * Move quantifiers tests * Move util tests * Fixed imports * Move TestArgQuantifiers to its correct place * Add correct packages to each test * Update tests ~README.md * Cleaning + renaming * Add base class * Use base class for util * Import the correct base class * Ignore bloop * Add baseSpec * Use new baseSpec * Change classes to use new baseSpec * Update BaseSpec.scala * Change CHANGES.md * Change CHANGES.md from `added` to `changed` * ops...

view details

George C. G. Barbosa

commit sha 91f0a753c70b1f61508062356a5d3431944a1799

Move images to docs/images (#84)

view details

George C. G. Barbosa

commit sha c29b1b1aa78a8198bea37ba1d8146bf0813746a2

Add link to odinson/actions to Scala CI badge (#85)

view details

danebell

commit sha 1988ae4afc0f751a3467c65078f734f005ef8a0c

Use default issue templates

view details

danebell

commit sha c1cf3d1144fd671e7bbebe100c50be8cb61c2967

Merge branch 'master' of github.com:lum-ai/odinson

view details

Becky Sharp

commit sha bb3e220ef954dc04bdd56ebeb2d5eed5c02cb207

variables as lists (#86) * var lists * Update CHANGES.md

view details

Becky Sharp

commit sha 63975e5259990ca78c44ec8839e65e6e277ec4e3

Made the indexing a method to be called by outside projects (#90) * Made the indexing a method to be called by outside projects * update CHANGES * comment about GenIterable

view details

Keith Alcock

commit sha 4f94544ebcc664f903412c3e93caadb2bc1a5e55

Implement a MentionFactory (#87) * Implement a MentionFactory * refactoring to simplfy method calls * renamed the namedcapture Co-authored-by: bsharpataz <bsharp@email.arizona.edu>

view details

bsharpataz

commit sha c4b60c955992e99d545e75d307fb9f82e98bc477

Merge branch 'master' into doc_draft

view details

bsharpataz

commit sha 5283502cb8d066b7bd4c3a45aafe74cf39f2d491

edits

view details

push time in 7 days

pull request commentlum-ai/odinson

Change tests for event rules

I love the idea of a TAG svg of all the docs!!!!! love love love it (at least all docs that are short enough to be feasible!) I think an id could be something like "ninjaTurtles" for the ninja turtles sentence, or "gummyBears", etc... it doesn't need to be exhaustive, but something human-readable that we can use as a reference

On Mon, Jun 22, 2020 at 7:18 PM George C. G. Barbosa < notifications@github.com> wrote:

External Email

@gcgbarbosa commented on this pull request.

In core/src/test/scala/ai/lum/odinson/events/EventSpec.scala https://github.com/lum-ai/odinson/pull/88#discussion_r443923495:

  •    val matchedForThisRole = groupedMatched(desiredRole)
    
  •    desired should have size matchedForThisRole.size
    
  •    for (d <- desired) {
    
  •      matchedForThisRole should contain (d)
    
  •    }
    
  •    // there shouldn't be any found arguments that we didn't want
    
  •    val unwantedArgs = groupedMatched.keySet.diff(groupedDesired.keySet)
    
  •    unwantedArgs shouldBe empty
    
  •  }
    
  • }
  • def createArgument(name: String, start: Int, end: Int): Argument = Argument(name, start, end)
  • def getJsonDocument(id: String): String = {
  • val json = Map(
  •  "1" -> """{"id":"48fb577b-f5ba-4e16-864f-f8ba20ba9cfa","metadata":[],"sentences":[{"numTokens":8,"fields":[{"$type":"ai.lum.odinson.TokensField","name":"raw","tokens":["The","consumption","of","gummy","bears","and","donuts","."],"store":true},{"$type":"ai.lum.odinson.TokensField","name":"word","tokens":["The","consumption","of","gummy","bears","and","donuts","."]},{"$type":"ai.lum.odinson.TokensField","name":"tag","tokens":["DT","NN","IN","NN","NNS","CC","NNS","."]},{"$type":"ai.lum.odinson.TokensField","name":"lemma","tokens":["the","consumption","of","gummy","bear","and","donut","."]},{"$type":"ai.lum.odinson.TokensField","name":"entity","tokens":["O","O","O","B-dessert","I-dessert","O","B-dessert","O"]},{"$type":"ai.lum.odinson.TokensField","name":"chunk","tokens":["B-NP","I-NP","B-PP","B-NP","I-NP","O","B-NP","O"]},{"$type":"ai.lum.odinson.GraphField","name":"dependencies","edges":[[1,0,"det"],[1,4,"nmod_of"],[1,7,"punct"],[4,2,"case"],[4,3,"compound"],[4,5,"cc"],[4,6,"conj"]],"roots":[1]}]}]}""",
    

Do you think using the id is enough?

I was planning to have a TAG svg of all documents used for test somewhere in a readme. Meanwhile the only thing I could think would be informative would be the document id.jk

— You are receiving this because your review was requested. Reply to this email directly, view it on GitHub https://github.com/lum-ai/odinson/pull/88#discussion_r443923495, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABJCPCOZ4G43SK5AM6EJT5TRYAGIBANCNFSM4OC73DZQ .

gcgbarbosa

comment created time in 15 days

delete branch lum-ai/odinson

delete branch : kwalcock-mentionFactory-bsharp

delete time in 15 days

PR opened lum-ai/odinson

Reviewers
renamed the namedcapture

Sorry!!!! one more thing, I like these names better!! (Sorry for the piecemeal, this time it's MY TURN to complain that you're too fast for me!!!)

+4 -4

0 comment

1 changed file

pr created time in 15 days

push eventlum-ai/odinson

bsharpataz

commit sha e3a1fa7b6a775dccbef86b03b14581e17d894e7f

renamed the namedcapture

view details

push time in 15 days

PR opened lum-ai/odinson

refactoring to simplfy method calls

First -- I love the idea of the MentionFactory, I think this will really simply things in external projects! I am doing this as a PR on your PR branch bc I think you do it that way in Eidos :)

@kwalcock -- met with @marcovzla and tried to refactor this a little to make it so that Mention doesn't need to care about what MentionFactory you're using when retrieving the arguments. Did this by (a) making the args a val, not a def, and (b) teaching the MentionFactory trait how to make Mention arguments. Likely this implementation will suffice most external needs, and won't often need to be overridden, but of course it can be.

If you like this, please merge into your branch and we can then merge into master?

+78 -37

0 comment

4 changed files

pr created time in 15 days

push eventlum-ai/odinson

bsharpataz

commit sha bc9f3ff8a7219843874301e3f56d75b87aa939f0

refactoring to simplfy method calls

view details

push time in 15 days

create barnchlum-ai/odinson

branch : kwalcock-mentionFactory-bsharp

created branch time in 15 days

push eventlum-ai/odinson

bsharpataz

commit sha 85150aae34c049a295cd559e73a13fbdfd21fdfd

comment about GenIterable

view details

push time in 15 days

pull request commentclulab/eidos

Clean up EidosMention

Yes -- it's DEF a step forward! re: how deduplication is performed, we try to pick the most complete/attachiest mention from a given span, but that bit of code is like our skeleton in the closet. you have made it more sound over the years FOR sure with tie breaking and all, but it's really complicated, and not that awesome, and we seem to still be continually dealing with issues with "duplicates". I guess I was wondering if we could do it fundamentally differently, knowing what we "know now" (i.e., 3 years into the program). Re: the subordinate question -- likely! I know you dig in and retrieve the subordinate ones, but that's after the fact, and I know that, for example, I had to be careful to dig in and manually propagate attachments to args etc., which felt gross... :)

On Fri, Jun 19, 2020 at 1:22 PM Keith Alcock notifications@github.com wrote:

External Email

Sure, but I would call it taking a step forward. In the past the "choice" was made largely by chance, so I don't mind changing it willfully. In the table above the left column is (except in maybe one case I discovered later) what Eidos was getting before I made the changes, so this is mostly a reversion to previous behavior. I'm not clear on how deduplication is performed by Odin. Does it ever happen that a mention found by Odin is made subordinate to another (as a trigger, argument, etc.) and then is no longer in the list which Odin would consider being a duplicate? By Odin I might mean inside an Eidos Finder or GlobalAction or something.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/clulab/eidos/pull/881#issuecomment-646847012, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABJCPCKOA2ZZK3U2PNVA2RDRXPCIFANCNFSM4OAKOB3A .

kwalcock

comment created time in 18 days

push eventlum-ai/odinson

bsharpataz

commit sha 2ab9126de47f56c43826f46873338bf68390b231

update CHANGES

view details

push time in 18 days

PR opened lum-ai/odinson

Reviewers
Made the indexing a method to be called by outside projects
+28 -22

0 comment

1 changed file

pr created time in 18 days

pull request commentlum-ai/odinson

Implement a MentionFactory

Yes, if enough people have seen it to conclude that it is a reasonable idea. It can be further refined.

Well I like it :)

kwalcock

comment created time in 18 days

create barnchlum-ai/odinson

branch : index_method

created branch time in 18 days

pull request commentclulab/eidos

Clean up EidosMention

hmm.... i worry about the choosing left vs right... something smells fishy to me. should we consider taking a step back and thinking about why we have so many duplicate mentions and if there's a better way to handle things?

On Fri, Jun 19, 2020 at 11:59 AM Keith Alcock notifications@github.com wrote:

External Email

Merged #881 https://github.com/clulab/eidos/pull/881 into master.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/clulab/eidos/pull/881#event-3463585893, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABJCPCPITCKITZCVJDOONI3RXOYSRANCNFSM4OAKOB3A .

kwalcock

comment created time in 18 days

issue commentml4ai/tomcat

Integrate multiplayer voice chat

  • [ ] survey, decide which tool to use (community of users)
  • [ ] read through docs / tutorials
  • [ ] get toy running, min working example
  • [ ] integrate with run_session
adarshp

comment created time in 19 days

issue commentlum-ai/odinson

ExtractorEngine.getTokens() should rely on config for determining raw field

unless I'm missing something, the call currently (in master) uses the displayField, which is filled in from the config and defaults to raw

myedibleenso

comment created time in 19 days

issue commentlum-ai/odinson

Support event-like mentions in state

Now that we have Event Mentions and progress on the State, we should revisit this @marcovzla @myedibleenso

myedibleenso

comment created time in 19 days

issue commentlum-ai/odinson

ExtractorEngine.getTokens() should rely on config for determining raw field

@myedibleenso I think this has been addressed by your displayField change. Should we close this?

myedibleenso

comment created time in 19 days

issue commentlum-ai/odinson

Documentation

@PeterAJansen note that #82 made some progress on the State -- look especially at the tests that changed. It's still WIP but wanted to ping you so you knew

PeterAJansen

comment created time in 19 days

pull request commentlum-ai/odinson

variables as lists

@marcovzla please check and let me know if this is what you're looking for. thanks!

BeckySharp

comment created time in 20 days

push eventlum-ai/odinson

Becky Sharp

commit sha e9299761a8f9de81f26cd650a2b3bd0fcad751b7

Update CHANGES.md

view details

push time in 20 days

MemberEvent

pull request commentlum-ai/odinson

Implement a MentionFactory

ok, cool, thanks!!

On Thu, Jun 18, 2020 at 8:53 AM Keith Alcock notifications@github.com wrote:

External Email

@kwalcock commented on this pull request.

In core/src/main/scala/ai/lum/odinson/Mention.scala https://github.com/lum-ai/odinson/pull/87#discussion_r442330541:

@@ -39,4 +39,15 @@ case class Mention( } }

  • def copy(mentionFactory: MentionFactory /= new DefaultMentionFactory/,

Probably not, if I understand the question correctly. This line in ExtractorEngine is where I intended customization to be incorporated. The engine gets a single factory when it is constructed which is hopefully good for the engine's entire lifespan. It might be slipped in elsewhere but this was the main point of attack.

def fromDirectory(config: Config, indexDir: Directory, mentionFactory: MentionFactory = defaultMentionFactory): ExtractorEngine = {

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/lum-ai/odinson/pull/87#discussion_r442330541, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABJCPCJPJAYHVZFZR5R3SM3RXI2AHANCNFSM4OBKS32Q .

kwalcock

comment created time in 20 days

Pull request review commentlum-ai/odinson

Implement a MentionFactory

 case class Mention(       }   } +  def copy(mentionFactory: MentionFactory /*= new DefaultMentionFactory*/,

@kwalcock do I understand correctly that this would allow for making a new "view" of the extraction with a different factory? If so -- seems like this may be my "entry point" for having a custom mentionfactory/mention in eidosdottir

kwalcock

comment created time in 20 days

pull request commentlum-ai/odinson

variables as lists

(ignore the and_lists of the branch name, I thought I would do something else but decided not to)

BeckySharp

comment created time in 20 days

PR opened lum-ai/odinson

Reviewers
variables as lists

We added this functionality to Odin, because it really helps with diffs etc (and seems like a reasonable representation). I'd like to port it here, if that's ok.

+41 -1

0 comment

2 changed files

pr created time in 20 days

create barnchlum-ai/odinson

branch : vars_and_label

created branch time in 20 days

push eventlum-ai/odinson

marcovzla

commit sha 1eda3005a98678676a6585e64de01e2feefc53f3

Mem (#59) * empty array * getAllMatches * empty array * refactor * fmt * deletes * changes

view details

marcovzla

commit sha 404fbe2a350a7b478d3e9c65a87ab6800c5e6b18

write compressed documents (#62) * write compressed documents * changes

view details

Dane Bell

commit sha 068a71bb1b85aae09898e9a8ff759be086b517cf

Update README with LREC paper

view details

marcovzla

commit sha 726fbebe23e3fb634fa9c9d4bf47ab2ebe4f9f96

Unicode normalization (#63) * unicode tests * upgrade common

view details

marcovzla

commit sha 3e7900c308c0a21312d2ab7df079ad502ab2cacf

Update CHANGES.md

view details

Becky Sharp

commit sha bba271d437b8e5e6a410f06ddc7d2c6d6b2e1943

Nullspans (#64) * handle null spans in args * handle nulls at diff levels * simplify span fix

view details

bsharpataz

commit sha f505026f46c38f5ae8637809e6ddb4f9780b7370

Merge branch 'master' into doc_draft

view details

Becky Sharp

commit sha b45d3adfc04e8730f6b7e3a7304b40416464499b

entrypoint (#67)

view details

Gus Hahn-Powell

commit sha b7d546dc25b3ae48e807819bf2cad2265863df18

Changes to REST API to support "event" matcher (#68) * Adjustments to Odinson REST API - Ability to retrieve OdinsonDocument JSON from filesystem - Revised OpenAPI schema (data model, etc.) * New endpoint for executing Odinson grammar - /api/search -> /api/execute/pattern - Removed /api/rich-search - /api/execute/grammar - Optionally takes a parentQuery - Updates to OpenAPI spec - GET -> PUSH * Cleanup

view details

Gus Hahn-Powell

commit sha dfdcb6dd6c50aac8dd2d8753fda732ad97a9dea9

Removed outdated config options (#69) Related to https://github.com/lum-ai/odinson/pull/68

view details

marcovzla

commit sha 705227bf413b06bdc7aad4876fa9b8d100e94257

upgrade common (#72)

view details

marcovzla

commit sha c3cf972a87537bbfaa1e89720b8fd26984f9d945

write buildinfo to index (#71) * write buildinfo to index * changes * removed data from buildinfo

view details

bsharpataz

commit sha 0188da978d775075572552aa4a6c8ebdbf3e8b46

start of github.io docs

view details

marcovzla

commit sha adb02c989d07e2a9f8c063ae0235bfb5982de722

all possible matches (#66) * compileRules * get all possible matches * api * api * bug * default args * changes * query() * mkQuery * deleted query() variants * tests * controller * rename `getAllPossibleMatches` to `disableMatchSelector` * changes

view details

marcovzla

commit sha 5b18761453b6c1cf79fe8d82cc8352a62a5b3086

rename argument (#73) * rename * rename

view details

bsharpataz

commit sha 2ff65b7fcff8b7decdb509e5ef32ae069f053fb2

Merge branch 'master' into doc_draft

view details

bsharpataz

commit sha af3e5054d985f35cfacf3acfa3ecb903051012f8

config for github pages

view details

bsharpataz

commit sha 83bbf67b8b195082b04dda9d4c04295f78af4ab5

wip

view details

bsharpataz

commit sha d0a0cd7b6fa7d7115c1c55c6be81fe506f1153b7

wip

view details

push time in 20 days

pull request commentclulab/eidos

Clean up EidosMention

love it! ❤️ ❤️ ❤️

kwalcock

comment created time in 20 days

pull request commentclulab/eidos

Clean up EidosMention

👍

kwalcock

comment created time in 21 days

pull request commentclulab/eidos

list view and a couple new

sorry @kwalcock I meant to add you as a reviewer, this is now ready :)

BeckySharp

comment created time in 22 days

push eventclulab/eidos

bsharpataz

commit sha 3132cb8de02ee5f7b996fb44140e2416b18ea5e9

triggers from other use case

view details

push time in 22 days

PR opened clulab/eidos

list view and a couple new

not readty -- wanted to see the tests

+189 -12

0 comment

1 changed file

pr created time in 22 days

create barnchclulab/eidos

branch : trigger_list

created branch time in 22 days

pull request commentlum-ai/odinson

**testing codcov**

Cool! hey -- when I looked at the report, I was able to see how many lines were "hit" and not, but not which lines were hit, is there a way to do that, maybe even with a diff view?

gcgbarbosa

comment created time in 23 days

issue openedlum-ai/odinson

Organization of Unit tests

Existing unit tests are mixed and shared resources are not shared.

We need to split out tests, consolidate resources for common use

created time in a month

issue openedlum-ai/odinson

Code coverage integration

In order to know how well Odinson is covered by tests, both the language and the system, we need to integrate some form of code coverage report.

Options include:

  • the scoverage sbt plugin
  • codecov
  • others

We should consider pricing/long-term support when making decisions

created time in a month

pull request commentclulab/asist-speech

Initial Code Migration

👍 yay!

jmculnan

comment created time in a month

MemberEvent
MemberEvent
MemberEvent
MemberEvent
MemberEvent

issue commentlum-ai/odinson

Documentation

@PeterAJansen Glad you're looking into Odinson, and if you have questions in the meantime of documentation, ask away, we're actively working on it. In addition to the links @myedibleenso posted, we have a readme in extra that describes some of the "setup" stuff (i.e., annotating to get the Odinson Documents, making the index, querying in the shell, etc.): https://github.com/lum-ai/odinson/tree/master/extra

PeterAJansen

comment created time in a month

push eventlum-ai/odinson

bsharpataz

commit sha 0188da978d775075572552aa4a6c8ebdbf3e8b46

start of github.io docs

view details

push time in a month

push eventmarcovzla/discobert

Becky Sharp

commit sha c67ea157533d48a1d59ea714d4f6348075bfc5ca

coarse labels and eval all

view details

push time in a month

push eventmarcovzla/discobert

Becky Sharp

commit sha d2b9ead1333dc6a8b139d5437292298105e3b0a9

merge function

view details

push time in a month

push eventmarcovzla/discobert

Becky Sharp

commit sha bb73f2f69583f7d82fac7c31e913f3564bd1d47f

train-test split

view details

push time in a month

push eventclulab/ConceptAlignment

bsharpataz

commit sha a9745858131ac1d824d22df16623402e3ae59daf

build.sbt and gitginore

view details

push time in a month

more