Lloyd Thompson

#AMMi ¦¦ #UnsupervisedLearning ¦¦ #Ai #ML #CV #Robotics🤖¦¦ #TreeClimber ¦¦ #Heresy

[Lloyd Thompson @nii4u]

@JoyNewsOnTV This is painful to watch and couldn't even finish it🤦🏾‍♂️🤦🏾‍♂️🤦🏾‍♂️

[Lloyd Thompson @nii4u]

@cybhoy @shattawalegh So now you guys are using Shatta's wall to promote Bin 💩💩 music smh. And you are still not ashamed???

[Edward Hu @edward_s_hu]

this discussion highlights a funny paradox in robot learning right now.
blind policies are pretty robust, but can't detect upcoming obstacles without bumping into them.
visual policies are often sensitive to noisy inputs, but they can sense upcoming obstacles.

[Fuse ODG @FuseODG]

One of its kind…This song allows you to explore all 54 Countries of Africa

To restore our voices and tell our stories accurately, we must first recognise the full spectrum of Africa’s reality…raa raa raaaahhhh 🔥🔥🔥🔥 #54Countries

https://t.co/DDsnp73Q9n

[Lloyd Thompson @nii4u]

#NewProfilePic

[Prof Lennart Nacke, PhD @acagamic]

Basic statistics concepts

(that all researchers should know)

[Lloyd Thompson @nii4u]

@aaditsh This is golden 🔥🔥🔥🔥

[Rishabh @Rixhabh__]

72+ AI tools to finish months of work in minutes.

1. Research

- ChatGPT
- Claude
- DeepSeek R1
- Gemini
- Abacus
- Perplexity

2. Image

- Midjourney
- Dall E-3
- Flux
- Stability AI
- Grok

3. CopyWriting

- Rytr
- Copy AI
- Writesonic
- Adcreative AI
- otio

4. Writing

- Jasper
- HIX AI
- Jenny AI
- Textblaze
- Quillbot

5. Website

- 10Web
- Durable
- Framer
- Style AI
- Landingsite

6. Video

- Sora
- Luma
- Kling
- Pika
- InVideo
- HeyGen
- Runway
- ImgCreator AI
- Morphstudio .xyz

7. Meeting

- Tldv
- Otter
- Noty AI
- Fireflies

8. SEO

- VidIQ
- Seona AI
- BlogSEO
- Keywrds ai
- Seona

9. Chatbot

- Droxy
- Chatbase
- Mutual info
- Chatsimple

10. Presentation

- Decktopus
- Slides AI
- Gamma AI
- Designs AI
- Beautiful AI
- PopAi

11. Automation

- Make
- Zapier
- Xembly
- Bardeen

12. UI/UX

- Figma
- Uizard
- UiMagic
- Photoshop

13. Design

- Canva
- Flair AI
- Clipdrop
- Autodraw
- Magician design

14. Logo Generator

- Looka
- Designs AI
- Brandmark
- Stockimg AI
- Namecheap

15. Audio

- Lovo ai
- Eleven labs
- Songburst AI
- Adobe Podcast

16. Startup

- Tome
- Ideas AI
- Namelix
- Pitchgrade
- Validator AI

17. Productivity

- Merlin
- Tinywow
- Notion AI
- Adobe Sensei
- Personal AI

18. Social media management

- Tapilo
- Typefully
- Hypefury
- TweetHunter

Follow me @Rixhabh__ for more.

[Jason Ma @JasonMa2020]

Introducing Dynamism v1 (DYNA-1) by @DynaRobotics – the first robot foundation model built for round-the-clock, high-throughput dexterous autonomy.

Here is a time-lapse video of our model autonomously folding 850+ napkins in a span of 24 hours with
• 99.4% success rate — zero human intervention
• 60% human throughput speed
• 4.3/5 quality ratings (set by the client)

A thread on our motivation, insights and results:

[Edward Hu @edward_s_hu]

introducing the belief state transformer: a new LLM training objective that learns (provably) rich representations for planning

bst objective is satisfyingly simple: just predict a "previous" token alongside the next token

come by our ICLR poster this thursday to chat!

[Sébastien Bratières @Seb_Bratieres]

At Translated we're sponsoring 5 language-related research projects, no ties attached, with 100k$ of grants every year. I'm looking forward to your application!

[Deep Learning Indaba @DeepIndaba]

Still unsure about how to apply for our 2025 annual event? 🤔✨ Join us for a live Q&A session where we'll walk you through the application process and answer all your questions!

📅 Monday, February 17, 2025
⏰ 6:30 PM GMT+1
📍 Online via Zoom platform
🎟️Registration link: https://t.co/sIhwVWYQCB

To make the most of the session, review our application guidelines in advance 👉https://t.co/oxxgBlqIT3 and come prepared with your questions –see you there! 🎤

Don't forget the deadline of the Indaba applications is on February 23rd, 2025.
#DLI2025
#Indaba2025

[Chelsea Finn @chelseabfinn]

Disappointed with your ICLR paper being rejected?

Ten years ago today, Sergey and I finished training some of the first end-to-end neutral nets for robot control 🤖

We submitted the paper to RSS on January 23, 2015.

It was rejected for being "incremental" and "unlikely to have much impact"

Our resubmission to NeurIPS was also rejected

It now has >4,000 citations (and more importantly, end-to-end training is widely accepted!)

It's also cool to think about what's changed and what's the same --
- The network was 92k parameters and trained on ~15 minutes of data
- The code was a combination of matlab, caffe, ROS, a custom CUDA kernel for speed, and a low-level 20 Hz controller in C++, all talking to each other. ROS+matlab was as bad as it sounds.
- We pre-trained the encoder and did inference off-board on a workstation with a larger GPU.
- We were paranoid about varying lighting messing up the network, so we did all the experiments after sunset (so long nights running experiments on the robot past 3 am)

Now, we have manipulation policies that are far more dextrous, far more generalizable, and maybe on the cusp of breaking into the real world. :)

(the paper: https://t.co/qDlGSdDExL)

[Lloyd Thompson @nii4u]

@tv3_ghana @tv3_ghana Paaa, please stop asking stupid questions. We have better issues to attend to.

[Paul Azunre @pazunre]

Great demo and review from @shaunnarmah of our recent TTS release ( @GhanaNLP @algorineinc ) Have you tried it yet?

Follow Shaunn’s channel to stay informed on major tech trends ���🏾 🙏🏾

We appreciate all the support we are getting recently from the community 🥰 Keep em coming!

[Paul Azunre @pazunre]

Don Quixote? a futile lonely thankless quest, destined for an early grave? 😂 💀 ☠️ Either way, I guarantee it is an interesting ride though 🙌🏾🤣🤪 @restofworld (link in first comment)

[Lloyd Thompson @nii4u]

@Alawa_depo @enokay69 Where is the time machine??

[Santiago @svpino]

LoRA is a genius idea.

To understand the fine-tuning of Large Language Models, you must understand how LoRA works.

By the end of this post, you'll know everything important about how it works.

Large Language Models are good generalists, but they have little specialization. We train them in many different tasks, so they know a bit about everything but not enough about anything.

Think of a kid who can play three different sports at a high level. While he can be proficient across the board, he won't get a scholarship unless he specializes. That's how the kid can reach his full potential.

We can do the same with these large models. We can train them to solve a particular task and nothing else.

We call this process "fine-tuning." We start with everything the model knows and adjust its knowledge to help it improve on the task we care about.

Fine-tuning is revolutionary, but it's not free.

Fine-tuning a large model takes time, care, and lots of money. Many companies can't afford the process. Some can't pay for the hardware. Some can't hire people who know how to do it. Most companies can't do either.

That's where LoRA comes in.

We realized we could approximate a large matrix of parameters using the product of two smaller matrices. There was a lot of wasted space within these large models. What would happen if we find a new, more optimal representation?

Did you ever buy a map at a gas station? Giant pages showing every small road, path, and lake around you. They were exhaustive but hard to navigate. These are like parameters in a large model.

LoRA turns a gas station map into a cartoon treasure map. Every useless parameter is gone. Only two roads, a palm tree, and a cross pointing at the treasure. We don't need to fine-tune the entire model anymore. We can only focus on the small treasure map that LoRA gives us.

It's a mind-blowing trick.

We can train the small approximation matrices from LoRA instead of fine-tuning the entire model. LoRA is cheaper, faster, and uses less memory and storage space.

You can also merge the approximation matrices with the model during deployment time. They work like simple adapters. You load up the one you need to solve a problem and use a different one for the next task.

Then, we have QLoRA, which makes the process much more efficient by adding 4-bit quantization. QLoRA deserves its own separate post.

The team at @monsterapis has created an efficient no-code LoRA/QLoRA-powered LLM fine-tuner.

What they do is pretty smart:

They automatically configure your GPU environment and fine-tuning pipeline for your specific model. For example, if you want to fine-tune Mixtral 8x7B on a smaller GPU, they will automatically use QLoRA to keep your costs down and prevent memory issues.

The @monsterapis platform specializes in no-code LoRA-powered fine-tuning. It's the fastest and most affordable offering for fine-tuning models in the market. They sponsored me and gave me 10,000 free credits for anyone who uses the code "SANTIAGO" in their dashboard:

https://t.co/DBsB9EXgYR

If you want to read their latest updates, get free credits and special offers, join their Discord server: https://t.co/oecoxEec14

TL;DR:

• Traditional fine-tuning trains the entire model. It requires a complex setup, higher memory, and expensive hardware.

• LoRA: Trains a small portion of the model. It's faster, requires much less memory, and affordable hardware.

• QLoRA: Much more efficient than LoRA, but it requires a more complex setup.

• No-code fine-tuning with LoRA/QLoRA: The best of both worlds. Low cost and easy setup.

[Paul Azunre @pazunre]

I’m going to @DevCongress’s upcoming Space. Will you join too? https://t.co/EZtNEt88XR