Ruan van der Merwe
@ruanvdmerwe1
Followers
125
Following
290
Media
15
Statuses
194
Technical Lead and Co-Founder | Cat Whisperer | https://t.co/95gYAXm3T0
Joined October 2015
RT @jeremyphoward: Bitcoin-turned-AI bros updating their profiles right now. ❤️ b̵i̵t̵c̵o̵i̵n̵ n̵f̵t̵ w̵e̵b̵3̵ A̵I̵ g̵e̵n̵e̵r̵a̵t̵i̵v̵….
0
188
0
❤️.
Is it just me? My timeline is only ❤️s.
0
0
0
Getting influenced to go back to my blink-182 days by @sama was not on my bingo card for 2023.
0
0
0
Highlight of the @DeepIndaba is easily building the transformers and LLM prac with these two! Thanks for the good time and memories @hereismari and @everlyn_asiko! Till next year :)
0
7
35
That moment when @NandoDF says your tranformers prac is amazing and recommended . @hereismari, Everlyn Chimoto
0
0
14
Sitting in the airport waiting for my flight to Accra and the @DeepIndaba fever is growing exponentially! If you are attending please reach out, would love to connect to as many people as possible this year. #deeplearningindaba.
1
1
12
A little unbiased opinion (😉) but the practicals are on another level this year. Looking forward to it!.
Have you had a chance to checkout all the exciting Practicals that will be happening at the Indaba? A huge shoutout goes out to all the teams that have been developing exciting code and content for all of us to learn from. Which one are you looking forward to the most?
0
1
4
RT @DeepIndaba: Have you had a chance to checkout all the exciting Practicals that will be happening at the Indaba? A huge shoutout goes ou….
0
24
0
Anybody found a way to quickly find the people I follow on twitter on threads?. Also, really liking threads! If your'e on it, you can find me at ruanvdmerwe15:
0
0
0
MAMLCon is an extension to model-agnostic meta learning (MAML), such that each inner learning loop—where a model “learns how to learn” new classes—ends with a single gradient update using one stored example from all the classes that the model has already seen.
0
0
0
Myself and Herman introduce a new method for handling continual learning tasks using meta-learning, called MAMLCon, which allows a network to be trained on new data while retaining knowledge from previous training iterations.
1
0
0
Super excited to say that our paper, "Mitigating Catastrophic Forgetting for Few-Shot Spoken Word Classification Through Meta-Learning," has been accepted at Interspeech 2023!. You can read the full paper here:
1
0
0
RT @ericsnowcrntly: PEP 684 has been accepted!!! Expect to see a per-interpreter GIL in Python 3.12. I'm so excited! This has been many….
0
229
0
Bit disappointed in the GPT-4 release. Defnitely a symptom of GPT tiredness, but here is to hoping the demo later makes me regret this tweet.
0
0
0
Like a famous shoe brand will say, just do it.
🇬🇭📣🇬🇭📣🇬🇭📣.Applications for the next DLI are now open! .We encourage applications from all members of the African machine learning community, in all career stages, whether you are in industry or academia. Apply by 6th March 2023: Tag your friends ‼️
0
1
4
I had a super fun time speaking at the @DeepIndabaX_ZA last year about everything transformer architecture related. If you want a deep understanding of the underlying architecture of Large Language Models like #ChatGPT, I hope this gets you started.
0
1
2
I have played The Last of Us 3 times, yet watching the @TheLastofUsTV opening chapter still managed to break me. Damn this story and it's brilaionous!.
0
0
0