Generalists vs. Specialists: Evaluating Large Language Models for Urdu
Let’s say you’re feeling a bit under the weather. You’ve got a scratchy throat, a nose that resembles a leaky tap, and a cough that rivals a 2-stroke rickshaw without a silencer. You start by calling up your GP (general practitioner) who, with their broad training of common ailments, reassures you that it’s probably just a mild cold – nothing that a little rest, some tea, and binge-watching a few seasons of your favourite series can’t fix. ...
UQA - Corpus for Urdu Question Answering
I think it was around 1999 when I first heard that Urdu is a low resource language. 25 years later Urdu is still considered low-resource despite having over 70 million native speakers. This is because large manually curated linguistic resources required for training are still not available for Urdu. One way around this barrier is to translate existing corpora from a resource-rich language (cough English cough) to Urdu. This seems like a chicken-and-egg problem though since good automatic translation systems require training resources. However, recent multimodal multitask models have reached a level where we can use their translations to set a baseline for more complex tasks like question answering. And we have done exactly that! ...
Concordance of Iqbal's Urdu Poetry | کشف الالفاظ - اردو کلامِ اقبال
tl;dr I built a KWIC concordance of Iqbal’s Urdu Poetry - Check out iqbal.chaoticity.com ! In distributional semantics, just like in interpretation of law, a word is known by the company it keeps. Observing how a word appears in context of other words allows us to identify usage nuances and gain deeper insights into the meaning and connotations of that particular term. Traditional word indexes do not show this context and thus miss out on the most remarkable ability of the human brain – recognising patterns. Which is perhaps why, in the late 50s, Hans Peter Luhn (the father of hash codes!) came up with the idea of creating such an index for technical documents that allowed users to see keywords in their original context, facilitating efficient information retrieval. This innovative approach marked a significant development in the field of text processing and laid the foundation for subsequent advancements in concordance-based analysis. He called this concept Keyword in Context (KWIC – pronounced quick) and it caught on … quickly. ...
#AcademicValentines
Roses are red Violets are blue Everyone’s equal When σ is μ
Almost eleven
This blog is almost 11 now. I’ll be trying out switching to an android client this year, just to see if it makes any difference in my posting frequency. Meanwhile, here’s a ’test swan'.
caffeine chronicles
Found this “graph” I created for #CraftyDataViz some time ago. It managed to get an honourable mention in the “Most Fun” category, so a much belated yay! :D (here’s the original tweet)
Of Mir and Meerkats
Breakfast Doodle
Accommodation in Cambridge
About 3-4 times a year, I get asked by students, visiting scholars, and people moving here for work about how to find accommodation in Cambridge. I usually reply by digging up the last few such emails and tweaking my response a little bit but it just crossed my mind that blogging about it might be more helpful. So here are some things I have learned about finding accommodation in Cambridge. I will not cover accommodation offered by colleges; if you are a student, you are probably better off looking at your college accommodation website. ...
Desk Descriptions
It’s been one week since I switched to a sit/stand desk setup. It’s time to report back. I’ve decided to try out a sit/stand setup for the next week. Will report back if it’s worth switching to permanently. A photo posted by Awais Athar (@awais_athar) on Aug 16, 2016 at 7:39am PDT Here are some of the points to note if you switch to a similar setup. ...
three sunsets
the same sun sets each day in different colours