Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

I put together a minimalistic pure-Rust, CPU-only implementation of the recent LFM2.5-8B-A1B language model from Liquid AI that you can directly link into your Rust projects😎

Maxime Chevalier

19,442 subscribers

14,034 views • 2 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Congrats to the Kimi.ai team! This is awesome. Great to see this level of research coming from open-source frontier model labs. I liked the paper so much I built a Rust implementation of it ;) Full AttnRes + Block AttnRes with two-phase inference, built using Burn (tensor library and Deep Learning Framework, in Rust, by Tracel AI). Runs on CPU, CUDA, Metal, wgpu. Includes an interactive TUI that trains a model live and visualizes depth attention evolving from uniform to selective in real time. Repo link and more on what is implemented in the comments.

Congrats to the Kimi.ai team! This is awesome. Great to see this level of research coming from open-source frontier model labs. I liked the paper so much I built a Rust implementation of it ;) Full AttnRes + Block AttnRes with two-phase inference, built using Burn (tensor library and Deep Learning Framework, in Rust, by Tracel AI). Runs on CPU, CUDA, Metal, wgpu. Includes an interactive TUI that trains a model live and visualizes depth attention evolving from uniform to selective in real time. Repo link and more on what is implemented in the comments.

abdel

95,651 views • 5 months ago

Why is Rust different than many/most programming languages? Alice Ryhl works on Google's Android Rust team, is a Rust language team advisor, and is a core maintainer of Tokio (the most widely-used async runtime in Rust) Timestamps: 00:00 Intro 04:09 Tokio: an overview 05:11 What Alice likes about Rust 12:48 Rust for TypeScript engineers 13:51 Moving from C++ to Rust 14:34 Memory safety 18:12 Garbage collection tradeoffs 21:46 Ownership, references, and borrowing 26:59 Unsafe in Rust 31:21 Crates and Cargo 35:55 Language design and RFCs 43:02 Building new features 46:30 Editions vs. versions 49:47 Getting paid to work on Rust 51:27 Contributing to Rust 53:03 Rust in the Linux kernel 55:45 AI use cases for Rust 1:01:35 Learning Rust 1:03:54 Book recommendation Brought to you by: • Antithesis – verify your system’s correctness without human review or traditional integration tests – and avoid bugs or outages. • Sentry – application monitoring software considered “not bad” by millions of developers Three things worth knowing about Rust: 1. Rust was designed to turn implicit failures into compile errors. Where other languages allow you to forget something, Rust makes an omission into a compilation error for things like null checks, uninitialized variables, or error propagation with the ‘?’ character. If you mess something up, it’s almost certain your program will not compile. If it does, at the very least you should see a lint warning. 2. Refactoring in Rust is safe and easy, thanks to the compiler. Alice: “I change a return type or struct field, then just fix the compiler errors until the compiler stops shouting. And then once I’ve done that, I’ve updated every place I need to update.” Rust’s focus on correctness makes refactoring it more straightforward than dynamically-typed languages and Java-style typed ones are to refactor. 3. “Editions” allow Rust to make breaking changes without ‘breaking’ anyone’s code. Rust editions (2015, 2018, 2021, 2024) can be mixed freely across crates. A library on the 2021 edition works seamlessly with a binary on the 2024 edition. This is how Rust evolves syntax (like adding async/await as keywords) without forcing an ecosystem-wide migration. Thanks a lot, Alice for this great discussion! And for your work on Rust.

Why is Rust different than many/most programming languages? Alice Ryhl works on Google's Android Rust team, is a Rust language team advisor, and is a core maintainer of Tokio (the most widely-used async runtime in Rust) Timestamps: 00:00 Intro 04:09 Tokio: an overview 05:11 What Alice likes about Rust 12:48 Rust for TypeScript engineers 13:51 Moving from C++ to Rust 14:34 Memory safety 18:12 Garbage collection tradeoffs 21:46 Ownership, references, and borrowing 26:59 Unsafe in Rust 31:21 Crates and Cargo 35:55 Language design and RFCs 43:02 Building new features 46:30 Editions vs. versions 49:47 Getting paid to work on Rust 51:27 Contributing to Rust 53:03 Rust in the Linux kernel 55:45 AI use cases for Rust 1:01:35 Learning Rust 1:03:54 Book recommendation Brought to you by: • Antithesis – verify your system’s correctness without human review or traditional integration tests – and avoid bugs or outages. • Sentry – application monitoring software considered “not bad” by millions of developers Three things worth knowing about Rust: 1. Rust was designed to turn implicit failures into compile errors. Where other languages allow you to forget something, Rust makes an omission into a compilation error for things like null checks, uninitialized variables, or error propagation with the ‘?’ character. If you mess something up, it’s almost certain your program will not compile. If it does, at the very least you should see a lint warning. 2. Refactoring in Rust is safe and easy, thanks to the compiler. Alice: “I change a return type or struct field, then just fix the compiler errors until the compiler stops shouting. And then once I’ve done that, I’ve updated every place I need to update.” Rust’s focus on correctness makes refactoring it more straightforward than dynamically-typed languages and Java-style typed ones are to refactor. 3. “Editions” allow Rust to make breaking changes without ‘breaking’ anyone’s code. Rust editions (2015, 2018, 2021, 2024) can be mixed freely across crates. A library on the 2021 edition works seamlessly with a binary on the 2024 edition. This is how Rust evolves syntax (like adding async/await as keywords) without forcing an ecosystem-wide migration. Thanks a lot, Alice for this great discussion! And for your work on Rust.

Gergely Orosz

51,961 views • 3 months ago

The Rustfall Outfitter just got a glow-up - Fresh new look & layout click any outfit to try it on - Instantly pay with RUST or CAPS - New limited-supply skins (More coming too) All rust that is generated from purchases of skins goes directly into the cashout pot. Only seem'd right that the first limited skin drop was toly 🇺🇸 The Grey! Only 20 available in the store once they're gone they're gone forever.

The Rustfall Outfitter just got a glow-up - Fresh new look & layout click any outfit to try it on - Instantly pay with RUST or CAPS - New limited-supply skins (More coming too) All rust that is generated from purchases of skins goes directly into the cashout pot. Only seem'd right that the first limited skin drop was toly 🇺🇸 The Grey! Only 20 available in the store once they're gone they're gone forever.

Rustfall

111,906 views • 1 month ago

I just "vibecoded" a Chess master (~2250 ELO) from scratch that runs locally on a Mac in Rust. I used to play chess semi-competitively, and I'm flabbergasted that you can just speak a 98% percentile chess engine into existence.

I just "vibecoded" a Chess master (~2250 ELO) from scratch that runs locally on a Mac in Rust. I used to play chess semi-competitively, and I'm flabbergasted that you can just speak a 98% percentile chess engine into existence.

Deedy

125,388 views • 5 months ago

AI coding agents can now extract any website's design system directly from the terminal. I love small projects like this that Hyperbrowser creates. Their repo is full of them.

AI coding agents can now extract any website's design system directly from the terminal. I love small projects like this that Hyperbrowser creates. Their repo is full of them.

AshutoshShrivastava

105,883 views • 2 months ago

Introducing YouTube video 🎥 link support in Google AI Studio and the Gemini API. You can now directly pass in a YouTube video and the model can usage its native video understanding capabilities to use that, with just a link! 🚢

Introducing YouTube video 🎥 link support in Google AI Studio and the Gemini API. You can now directly pass in a YouTube video and the model can usage its native video understanding capabilities to use that, with just a link! 🚢

Logan Kilpatrick

792,505 views • 1 year ago

NEW: LiquidAI just released LFM2.5-350M, a tiny model that brings agentic AI and tool-calling capabilities to resource-constrained environments. 🤯 It can even run locally in your browser via WebGPU, serving as a powerful companion while you browse the web. Try the demo! 👇

NEW: LiquidAI just released LFM2.5-350M, a tiny model that brings agentic AI and tool-calling capabilities to resource-constrained environments. 🤯 It can even run locally in your browser via WebGPU, serving as a powerful companion while you browse the web. Try the demo! 👇

Xenova

26,267 views • 4 months ago

Hear directly from Google’s VP of Product for Gmail, Blake Barnes, on how we’re turning your inbox into a personal, proactive assistant. From instant answers with AI Overviews to AI Inbox that surfaces your top priorities. Learn how the new era of Gmail can help you focus on what matters most.

Hear directly from Google’s VP of Product for Gmail, Blake Barnes, on how we’re turning your inbox into a personal, proactive assistant. From instant answers with AI Overviews to AI Inbox that surfaces your top priorities. Learn how the new era of Gmail can help you focus on what matters most.

Gmail

20,316 views • 7 months ago

I'm building an app from the ground up, and I'm filming every part of it. It's called Slopwatch. It's a great way to pick up tips for using AI Coding tools like a real engineer. First step is choosing the language. Rust/Go? Node/Bun? Only way to find out is to /grill-me.

I'm building an app from the ground up, and I'm filming every part of it. It's called Slopwatch. It's a great way to pick up tips for using AI Coding tools like a real engineer. First step is choosing the language. Rust/Go? Node/Bun? Only way to find out is to /grill-me.

Matt Pocock

57,236 views • 4 months ago

You can now automate any task on your phone by letting AI control it AutoGLM from Zai is a 100% open source vision-language model that: - Understands what's on your screen - Acts autonomously from a prompt - Totally private (works LOCALLY) Tutorial ↓

You can now automate any task on your phone by letting AI control it AutoGLM from Zai is a 100% open source vision-language model that: - Understands what's on your screen - Acts autonomously from a prompt - Totally private (works LOCALLY) Tutorial ↓

Paul Couvert

169,881 views • 8 months ago

Most C Programmers don't have a clue as to how the CPU works. Here is a lecture from my course on the C Language that introduces how to visualise the CPU as a C (or general software) Programmer.... More Details are at: Being able to model the behaviour of CPU provides a solid base to reason about your code. The summary is - - The CPU is a state machine which maintains it's states in registers. - It has busses to memory for requesting data/instruction read/writes. - Interrupts to attend to events. I discuss this and more in my course on the C Language.

Most C Programmers don't have a clue as to how the CPU works. Here is a lecture from my course on the C Language that introduces how to visualise the CPU as a C (or general software) Programmer.... More Details are at: Being able to model the behaviour of CPU provides a solid base to reason about your code. The summary is - - The CPU is a state machine which maintains it's states in registers. - It has busses to memory for requesting data/instruction read/writes. - Interrupts to attend to events. I discuss this and more in my course on the C Language.

Piyush Itankar

28,019 views • 1 year ago

New LFM2.5-2.6B hits DeepSeek-V4 level on tool calling and runs 3.7x faster! We ran Liquid AI 's new LFM2.5-2.6B against DeepSeek-V4-Flash on one box with 4x RTX 5090. Both got the same three jobs, and each one only completes if the model fires every tool call Topics: -weather and local time in six cities -one budget into six currencies -four hotels checked and booked for one date Outputs: LFM2.5-2.6B: 35/35 tool-calls, 366 tok/s, 19s DeepSeek-V4-Flash: 35/35 tool-calls, 77 tok/s, 70s Both models made all 35 calls, and both fired twelve of them in one turn. LFM was trained for agents, and tool calling is where that shows. Strong result for a model you can run right on your phone

New LFM2.5-2.6B hits DeepSeek-V4 level on tool calling and runs 3.7x faster! We ran Liquid AI 's new LFM2.5-2.6B against DeepSeek-V4-Flash on one box with 4x RTX 5090. Both got the same three jobs, and each one only completes if the model fires every tool call Topics: -weather and local time in six cities -one budget into six currencies -four hotels checked and booked for one date Outputs: LFM2.5-2.6B: 35/35 tool-calls, 366 tok/s, 19s DeepSeek-V4-Flash: 35/35 tool-calls, 77 tok/s, 70s Both models made all 35 calls, and both fired twelve of them in one turn. LFM was trained for agents, and tool calling is where that shows. Strong result for a model you can run right on your phone

atomic.chat

110,480 views • 12 days ago

Standing together in support of Venezuela. ❤️ A heartfelt thank you to everyone who contributed at the park today—your support helps bring essential aid to those suffering from the recent earthquake. You can still make a donation through the link:

Standing together in support of Venezuela. ❤️ A heartfelt thank you to everyone who contributed at the park today—your support helps bring essential aid to those suffering from the recent earthquake. You can still make a donation through the link:

David Ortiz

100,894 views • 1 month ago

In Rust, error handling is opt-out, not opt-in. Alice Ryhl (Rust for Android at Google, Rust language advisor & Tokio maintainer) explains: “The other thing I think is quite good is error handling. So on one hand, Rust doesn't really use exceptions, so it actually returns the error as a value. So you return a value that's, using an enum, either the result or the error. And the way this is done is that there's an operator `?` - so you write: my_function and then a question mark at the end - which means if this function fails, return the error. So it's really easy to handle errors but it's not serial characters like it is with exceptions, so it's explicit. On the other hand, if you forget to put the question mark, that's a compilation error. So you have to check it, and of course you can also have it manually, but the point is it's this idea of there are these things where you write some code and there's some implicit error condition you didn't think of and now you just took down your server or something.”

In Rust, error handling is opt-out, not opt-in. Alice Ryhl (Rust for Android at Google, Rust language advisor & Tokio maintainer) explains: “The other thing I think is quite good is error handling. So on one hand, Rust doesn't really use exceptions, so it actually returns the error as a value. So you return a value that's, using an enum, either the result or the error. And the way this is done is that there's an operator `?` - so you write: my_function and then a question mark at the end - which means if this function fails, return the error. So it's really easy to handle errors but it's not serial characters like it is with exceptions, so it's explicit. On the other hand, if you forget to put the question mark, that's a compilation error. So you have to check it, and of course you can also have it manually, but the point is it's this idea of there are these things where you write some code and there's some implicit error condition you didn't think of and now you just took down your server or something.”

The Pragmatic Engineer

52,059 views • 2 months ago

🛝🥏🏃‍♀️ Have you checked out the new AI-powered Builder agent on our homepage? Use natural language inputs to create maps for travel, business, special projects and anything else you can think of - limited only by your imagination ➡️

🛝🥏🏃‍♀️ Have you checked out the new AI-powered Builder agent on our homepage? Use natural language inputs to create maps for travel, business, special projects and anything else you can think of - limited only by your imagination ➡️

Google Maps Platform

63,805 views • 8 months ago

I put together a list of over 90 things that are different in the new 2026 Model Y vs the legacy Model Y. This is my video going over everything. 4K YouTube link in the thread below if you want to watch it there.

I put together a list of over 90 things that are different in the new 2026 Model Y vs the legacy Model Y. This is my video going over everything. 4K YouTube link in the thread below if you want to watch it there.

Sawyer Merritt

275,354 views • 1 year ago

I decided to put together all my MCP posts in a single PDF. It covers: - The fundamentals of MCP - Explanations with visuals and code - 11 hands-on projects for AI engineers Download link in next tweet!

I decided to put together all my MCP posts in a single PDF. It covers: - The fundamentals of MCP - Explanations with visuals and code - 11 hands-on projects for AI engineers Download link in next tweet!

Akshay 🚀

218,658 views • 1 year ago

Think of Model Context Protocol as a USB port for your AI. 🔌 You can plug in documentation, tools, and context directly into Copilot CLI. Scott Hanselman 🌮 shows how to set it all up. 👇

Think of Model Context Protocol as a USB port for your AI. 🔌 You can plug in documentation, tools, and context directly into Copilot CLI. Scott Hanselman 🌮 shows how to set it all up. 👇

GitHub

44,363 views • 6 months ago

Okay it works!!!! Now making the interface for it, I named it [ 👋 Hold a product ] You upload your product photo, you take a photo and it generates one with your AI model holding it Then from there you can press [ Talking video ], write a script and your AI model will present the product for you The AI character is from @lucataco93's tweets 😊

Okay it works!!!! Now making the interface for it, I named it [ 👋 Hold a product ] You upload your product photo, you take a photo and it generates one with your AI model holding it Then from there you can press [ Talking video ], write a script and your AI model will present the product for you The AI character is from @lucataco93's tweets 😊

@levelsio

188,528 views • 1 year ago