[01:09:36] 10Scoring-platform-team, 10Discovery-Search, 10Growth Design, 10Growth-Team (Current Sprint), 10Patch-For-Review: Newcomer tasks: UX changes for ORES topics - https://phabricator.wikimedia.org/T244421 (10Catrope) The main part of this (the new UI for ORES topics) is now in code review. Next, I'm going to... [01:19:51] 10MediaWiki-extensions-ORES, 10Scoring-platform-team, 10Discovery-Search, 10NewcomerTasks 1.1, and 2 others: Expose ORES drafttopic data in ElasticSearch via a custom CirrusSearch keyword - https://phabricator.wikimedia.org/T240559 (10Tgr) This is done. The functionality is not available on all wikis since... [01:26:57] 10Scoring-platform-team, 10Discovery-Search, 10Growth Design, 10Growth-Team (Current Sprint), 10Patch-For-Review: Newcomer tasks: UX changes for ORES topics - https://phabricator.wikimedia.org/T244421 (10Tgr) The section order is configurable at the bottom of [[https://www.mediawiki.org/wiki/MediaWiki:Ne... [03:26:46] 10Scoring-platform-team, 10drafttopic-modeling, 10Documentation, 10good first task: Add topic information to the ores-support-checklist - https://phabricator.wikimedia.org/T245068 (10srishakatux) @Chtnnh Hi! This task is available to be claimed. If you would like to work on it, feel free to assign it to yo... [07:50:32] 10Scoring-platform-team, 10drafttopic-modeling, 10Documentation, 10good first task: Add topic information to the ores-support-checklist - https://phabricator.wikimedia.org/T245068 (10Chtnnh) a:03Chtnnh Thank you @srishakatux. I will reach out here in case I hit a roadblock. Thanks for the support. [11:59:34] 10Scoring-platform-team, 10drafttopic-modeling, 10Documentation, 10good first task: Add topic information to the ores-support-checklist - https://phabricator.wikimedia.org/T245068 (10Chtnnh) Hey @srishakatux! So I got the app running on my local system , figured out where exactly to make the changes and... [13:25:56] 10Scoring-platform-team, 10articlequality-modeling, 10artificial-intelligence: Add text complexity scoring to article quality models - https://phabricator.wikimedia.org/T246438 (10Halfak) [13:27:16] 10Scoring-platform-team, 10artificial-intelligence, 10good first task: Text complexity scoring - https://phabricator.wikimedia.org/T155843 (10Halfak) Hello @Chtnnh! We've had a group of newcomers take this task on once before. We were able to get their work deployed for a period of time, but I think we nee... [13:27:26] 10Scoring-platform-team, 10artificial-intelligence: Text complexity scoring - https://phabricator.wikimedia.org/T155843 (10Halfak) [14:06:18] 10Scoring-platform-team, 10artificial-intelligence: Text complexity scoring - https://phabricator.wikimedia.org/T155843 (10Chtnnh) @Halfak What happened to the solutions deployed before? Were they ineffective in solving the issue or were there other constraints? The AI posed in the original question could be... [14:07:44] 10Scoring-platform-team, 10articlequality-modeling, 10artificial-intelligence: Add text complexity scoring to article quality models - https://phabricator.wikimedia.org/T246438 (10Chtnnh) a:03Chtnnh @Halfak It might take me some time to get a familiarity with the wikimedia code already used here, but I am... [14:21:22] 10Scoring-platform-team, 10artificial-intelligence: Text complexity scoring - https://phabricator.wikimedia.org/T155843 (10Halfak) The biggest limitation of the solution deployed previously is that it didn't have a clear use-case. E.g. let's say that an article scores an 8.3 flesch reading ease. What does th... [14:22:13] 10Scoring-platform-team, 10articlequality-modeling, 10artificial-intelligence: Add text complexity scoring to article quality models - https://phabricator.wikimedia.org/T246438 (10Halfak) Sounds great! If you do IRC, please join us in `#wikimedia-ai` on Freenode. Our team and the volunteers who work with u... [15:05:29] @hal [15:05:29] 04Error: Command “hal” not recognized. Please review and correct what you’ve written. [15:05:52] 10ORES, 10Scoring-platform-team (Current), 10Operations: ORES uwsgi consumes a large amount of memory and CPU when shutting down (as part of a restart) - https://phabricator.wikimedia.org/T242705 (10Halfak) I ran a test with gunicorn on my laptop and I was able to replicate this behavior. I started the se... [15:07:42] Hi chtnnh! Welcome! [15:07:51] Hi Halfak! [15:08:20] Hello all, I am chtnnh, new to both wikimedia and open source [15:12:23] I'm online early today. Usually kbazira and accraze are around. They are engineers on my team. They'd be happy to help if you don't see me around. [15:12:41] Otherwise, most of the people in this chat room have an interest or specific projects around AI :) [15:12:55] This is a great place to ask questions. [15:13:09] Thats a relief [15:13:15] If you say "halfak", I'll get a ping so it's a good way to get my attention if you send a message and I'm not responding. [15:13:19] Because I feel quite lost in the code here [15:13:32] sure, I will keep that in mind [15:13:42] Ha. Yeah, it's a bit of a jump to work through the dependency injection framework at first. [15:13:50] exactly [15:13:53] But I promise that it makes life easier later. [15:14:06] yeah I thought about it and saw why [15:14:22] do you have five minutes halfak? I have some minor questions [15:20:08] Sure! Ask away [15:21:43] 10ORES, 10Scoring-platform-team (Current), 10Operations: ORES uwsgi consumes a large amount of memory and CPU when shutting down (as part of a restart) - https://phabricator.wikimedia.org/T242705 (10Halfak) I was able to replicate the behavior with a very simple Flask application. app.py: ` $ cat app.py... [15:21:47] how do I build the article quality models and test them locally [15:22:08] What operating system are you running? [15:22:13] ubuntu 19 [15:22:35] Awesome! That'll be helpful. [15:22:37] https://github.com/wikimedia/articlequality/blob/master/Makefile [15:23:21] great, I will add the changes you have suggested to the feature extraction [15:23:27] We use the Makefile to implement a model building pipeline. If you have all of the requirements installed, you should be able to do something like "make models/enwiki.nettrom_wp10.gradient_boosting.model" and watch the whole process come together. [15:23:38] and then test the models locally to see if there is improvement [15:23:49] right. [15:23:53] aha sounds good [15:24:23] trivial question: [15:24:36] add the changes in a new branch to enwiki.py [15:24:40] test model [15:24:44] anything im missing? [15:26:05] what do you want me to do about breaking the text into sections in the revscoring parsed.py file? [15:26:18] simply call the get_sections() from the moodule? [15:33:57] I think we'll need to start by making a PR to revscoring to add the ability to split the article into sections. [15:34:23] yeah ofcourse that makes more sense [15:34:42] let me work on that [15:36:31] 10Scoring-platform-team, 10articlequality-modeling, 10artificial-intelligence: Add text complexity scoring to article quality models - https://phabricator.wikimedia.org/T246438 (10Halfak) Here's an example of how you can manually run a "parsed wikitext" datasource from the revscoring repository: ` $ python... [15:36:47] ^ chtnnh [15:36:58] I just posted an example for how you can experiment with running this stuff. [15:37:14] I used the "templates" datasource but you'll be adding a "sections" datasource right next to that so it should work similarly. [15:38:48] okay [15:39:07] so what this model does is solve the template in the given form [15:39:16] Foo bar {{}} .. etc [15:39:18] right? [15:43:24] right [15:52:09] okay I feel like im getting a hang of this now xD [15:52:14] thanks so much halfak [15:52:23] \o/ [15:52:25] Awesome. [15:52:27] :) [16:03:26] 10Scoring-platform-team, 10artificial-intelligence: Text complexity scoring - https://phabricator.wikimedia.org/T155843 (10Chtnnh) Right, so what we are doing in T246438 is essentially incorporating complexity into article quality, leaving this task redundant. [16:36:03] 10ORES, 10Scoring-platform-team (Current), 10Operations: ORES uwsgi consumes a large amount of memory and CPU when shutting down (as part of a restart) - https://phabricator.wikimedia.org/T242705 (10Halfak) @ACraze and I were discussing this and we were wondering if maybe it's time to try threads. In some s... [16:49:40] 10ORES, 10Scoring-platform-team (Current), 10Operations: ORES uwsgi consumes a large amount of memory and CPU when shutting down (as part of a restart) - https://phabricator.wikimedia.org/T242705 (10Halfak) We also talked about https://medium.com/@pgjones/quart-a-asyncio-alternative-to-flask-32666ae2abb0 - h... [17:00:38] accraze halfak do i define a class or method in revscoring parsed.py to implement splitting of articles into sections? [17:01:31] as in, there are three classes in the file: execute_method, get_key and Revision [17:01:50] I think a method would be a better option but am unsure as to which classs to put it under [17:02:34] chtnnh, you can put it into the parsed.py file. I think defining a method you use in the Datasource makes sense. [17:03:02] which class would be best to define it under? [17:03:20] oh sorry my bad [17:03:52] Look at how self.templates is defined and try to do something like that. [17:04:15] alright, you got it [19:43:40] halfak here's some good info re: gunicorn/threads/asyncio etc [19:43:44] https://medium.com/building-the-system/gunicorn-3-means-of-concurrency-efbb547674b7 [19:44:59] 10Scoring-platform-team, 10drafttopic-modeling, 10Documentation, 10good first task: Add topic information to the ores-support-checklist - https://phabricator.wikimedia.org/T245068 (10Chtnnh) {F31649291} Please check out the attached screenshot to see if the changes seem fine. I have changed the `max-width... [19:47:50] accraze, I tried the psuedo-threads approach and it seemed to work just fine. [19:48:25] awesome! was just going to say psuedo-threads looked promising [19:52:07] 10Scoring-platform-team, 10ORES-Support-Checklist: Document and share operational details of ores-support-checklist - https://phabricator.wikimedia.org/T222271 (10Chtnnh) Hello @Halfak! Can you elaborate on the further action required on this task? Thanks, Chaitanya [19:56:28] 10Scoring-platform-team, 10drafttopic-modeling, 10Documentation, 10good first task: Add topic information to the ores-support-checklist - https://phabricator.wikimedia.org/T245068 (10Halfak) It looks like the check marks aren't showing up for the topic models for arwiki, cswiki, and enwiki. Otherwise, I l... [19:58:49] 10Scoring-platform-team, 10drafttopic-modeling, 10Documentation, 10good first task: Add topic information to the ores-support-checklist - https://phabricator.wikimedia.org/T245068 (10Chtnnh) Because the data its sourcing was last updated in 2018. I had the same doubts at first. Seems to check out the story... [20:07:45] accraze, I'm looking for a good thing to link people to on beta so they can look at Jade. [20:07:48] Any suggestions? [20:08:21] It looks like https://en.wikipedia.beta.wmflabs.org/wiki/Jade:Diff/376901 is running old cold. [20:08:23] *code [20:14:33] whoa yeah that is a completely different schema [20:14:36] 10Jade, 10Scoring-platform-team (Current), 10Operations, 10TechCom, and 4 others: Deploy pilot of Jade to a small set of wikis. - https://phabricator.wikimedia.org/T183381 (10Halfak) [20:16:30] lemme make a new jade page real quick [20:21:10] here's a new link halfak: https://en.wikipedia.beta.wmflabs.org/wiki/Jade:Diff/1234556 [20:22:29] weird it looks like the assets only load if you are logged in ...? [20:28:10] the issue doesn't seem to happen on deployment: https://deployment.wikimedia.beta.wmflabs.org/wiki/Jade:Diff/1337 [20:30:33] 10Jade, 10Scoring-platform-team, 10CommRel-Specialists-Support: Design Jade pilot deployment plan with the Scoring Platform team - https://phabricator.wikimedia.org/T246486 (10Halfak) [20:31:11] Oh... What is this deployment.wikimedia... wiki? I've never seen this before! [20:32:12] Eh. Either way, I think I'm done for now. I got a support request into comm rel. After writing the request I think it is clear that I need to polish our user-facing docs so I think I'll be focusing on that next week. [20:32:25] Have a good weekend accraze and chtnnh! [20:32:54] you too halfak! [20:33:00] later halfak [20:34:01] 10Scoring-platform-team, 10articlequality-modeling, 10artificial-intelligence: Add text complexity scoring to article quality models - https://phabricator.wikimedia.org/T246438 (10Chtnnh) {F31649416} I have double checked the path, but the error persists. The `get_sections()` method is specified in the Wi... [20:39:53] halfak: i was apparently dropped from membership of github.com/wikimedia because of 2FA, can you re-add me whenever you find time? [20:57:15] 10Scoring-platform-team, 10articlequality-modeling, 10artificial-intelligence: Add text complexity scoring to article quality models - https://phabricator.wikimedia.org/T246438 (10Sumit) >>! In T246438#5927946, @Chtnnh wrote: > {F31649416} > > I have double checked the path, but the error persists. > > Th...