Ph.D. / NLP Researcher | AI Engineer @Towards_AI | Link to my monthly newsletter about NLP trends and blog👇
Joined July 2021
- Tweets1.3K
- Following1.1K
- Followers1.4K
- Likes3.3K
Ala Falaki retweeted
Kimi K3 goes open weights tomorrow
Heres what to do to prepare:
1) Find $500,000 to spend on hardware
Ala Falaki retweeted
Conversation with an engineer friend:
> me: open source models are so good
> him: yeah glm5.2 is great
> me: do you use them
> him: no - do you
> me: no
Ala Falaki retweeted
(1) Today we're releasing Muse Spark 1.1 -- a strong agentic and coding model at a very low price. It's available through our new Meta Model API and in Meta AI.
Ala Falaki retweeted
An amazing blog post dropped on @huggingface explaining how today's LLM inference engines like @vllm_project work!
The concept of "continuous batching" is explained, along with KV-caching, attention masking, chunked prefill, and decoding.
Continuous batching is the idea of concatenating prompts and cleverly masking the attention to avoid wasting compute on padding tokens
1/2
Ala Falaki retweeted
🚨 Product Hunt Launch in 4 Days!
Nixie Social is already helping small businesses automate their social content, now we’re bringing it to the Product Hunt community.
🎁 PH-only deal: 30% off with PH30OFF
📅 Launching in 4 days!
#ProductHunt #StartupLaunch #AItools
Ala Falaki retweeted
The EU saw someone out-regulating them and was NOT having it
TikTok could be banned in Europe as well as the US 9to5mac.com/2024/04/30/tikto… by @benlovejoy
Ala Falaki retweeted
🥁 Llama3 is out 🥁
8B and 70B models available today.
8k context length.
Trained with 15 trillion tokens on a custom-built 24k GPU cluster.
Great performance on various benchmarks, with Llam3-8B doing better than Llama2-70B in some cases.
More versions are coming over the next few months.
llama.meta.com/llama3/
Ala Falaki retweeted
Tweeteasy is live on ProductHunt Today! 🚀
I've spent 5 months building this tool to help myself and other creators tweet more easily and grow faster.
This is my first launch on PH. I really need your support on this special day 🥹
Retweet it, very, very much appreciated! 🙏
producthunt.com/posts/tweete…
Ala Falaki retweeted
Our service was recently featured on the BizX Magazine!
Discover how Nixie Social is revolutionizing social media management in the January issue! bizxmagazine.com/biz-x-magaz… (Page 12)
Special thanks to @BizXmagazine for covering our story.
#SocialMediaManagement
Ala Falaki retweeted
Just launched: Advanced Retrieval Augmented Generation course with @langchain & @llama_index!
40+ free lessons and practical projects on building RAG apps for production by @activeloop, @towards_AI, @intel Disruptor, & @llama_index. What's in store?
Ala Falaki retweeted
OpenAI's security team noticed that a group reverse engineered and was abusing ChatGPT's internal API. Instead of shutting them down, they quickly replaced ChatGPT with CatGPT… and then lurked in the attackers' Discord to watch the chaos. Absolute legend. youtube.com/watch?v=PeKMEXUr…
ALT CatGPT
Ala Falaki retweeted
Excited to announce a new @huggingface space to help with one of machine learning's biggest questions:
How much space does {X} model take in vRAM? And most importantly: when using `device_map="auto"`
huggingface.co/spaces/hf-acc…
Ala Falaki retweeted
We wish you a wonderful week #buildinpublic !!!
Don't forget, that we are having a "classic" Build, Show, & Tell this week on Wednesday at 7PM CEST!
🚨 Build, show, & tell "classic" is back! (#12) 🚨
Next week, @NLPiation is showing @_Socialization which is a a platform that delivers tailored social media content for businesses using AI.
Let's give @NLPiation a friendly roast! 😎
📅 Oct 18 🕖 7PM CEST
#buildinpublic
Ala Falaki retweeted
Launching a Free Course on Training & Fine-Tuning Large Language Models (LLMs) for Production in collaboration with @towards_AI and @intel Disruptor! This course was also created with invaluable support from our friends @cohere, @LambdaAPI, & @wandb.
Ala Falaki retweeted
Does a language model trained on “A is B” generalize to “B is A”?
E.g. When trained only on “George Washington was the first US president”, can models automatically answer “Who was the first US president?”
Our new paper shows they cannot!
Ala Falaki retweeted
📝The paper introduces a method for integrating the dimensionality reduction capacity of autoencoders (AE) into a #Transformer-based networks, aiming to effectively reduce the model's overall size. @NLPiation
🔗View details here: mdpi.com/2504-4990/5/3/45
Ala Falaki retweeted
#NeurIPS2023 A reviewer just wrote “thanks for you explanation” to our 600-word rebuttal addressing all the questions he/she asked. Exactly these four words. Just that simple. I’m in shock. Who am I speaking to?
