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.

My next video will be positively awesome! - Train sub-nano 0.1B language models on custom narrow tasks - Write model SDKs/APIs that run blazingly fast in client machine. I'm talking 350 tok/s with 0.3GB peak memory - How to create synthetic datasets and ship vertical SLMs

AVB

10,988 subscribers

64,720 views • 3 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

Watch this 45 min video to learn how to create synthetic datasets and train tiny (100M params) local language models that expertise on narrow tasks. Code, datasets, models, harnesses all in comments.

Watch this 45 min video to learn how to create synthetic datasets and train tiny (100M params) local language models that expertise on narrow tasks. Code, datasets, models, harnesses all in comments.

AVB

35,576 views • 1 month ago

Create datasets, run evals, and even train models directly in Cursor with the Hugging Face plugin. Here's Ben Burtenshaw to show you how:

Create datasets, run evals, and even train models directly in Cursor with the Hugging Face plugin. Here's Ben Burtenshaw to show you how:

edwin

17,119 views • 4 months ago

Apple built a large foundation model and fine-tuned it on multiple tasks. But they are doing something very clever: They load a single model in memory and use different adapters to specialize the model on the fly. I recorded a video to show you how to write the code to do the same thing Apple is doing. I explain everything step by step. Here is what I'll show you in the video: 1. We'll load two datasets 2. Then load a large model 3. Then, we'll fine-tune the model on both datasets I'll use LoRA to fine-tune the model. This process creates two small adapters, each specializing in solving one of the datasets. The base model's original parameters will remain unchanged. From here: 4. We'll generate a list of tasks 5. We'll load the correct adapter to solve each task The large model I'm using needs 346 MB of memory, but I only need to load it once. Each adapter is only 2.7 MB. I only need to load the base model once and pair it with any of the fine-tuned adapters. Minimum memory footprint and I can solve multiple tasks. Hope this helps!

Apple built a large foundation model and fine-tuned it on multiple tasks. But they are doing something very clever: They load a single model in memory and use different adapters to specialize the model on the fly. I recorded a video to show you how to write the code to do the same thing Apple is doing. I explain everything step by step. Here is what I'll show you in the video: 1. We'll load two datasets 2. Then load a large model 3. Then, we'll fine-tune the model on both datasets I'll use LoRA to fine-tune the model. This process creates two small adapters, each specializing in solving one of the datasets. The base model's original parameters will remain unchanged. From here: 4. We'll generate a list of tasks 5. We'll load the correct adapter to solve each task The large model I'm using needs 346 MB of memory, but I only need to load it once. Each adapter is only 2.7 MB. I only need to load the base model once and pair it with any of the fine-tuned adapters. Minimum memory footprint and I can solve multiple tasks. Hope this helps!

Santiago

84,747 views • 1 year ago

finetuning LLMs will never be the same! Unsloth just launched an open-source web UI to run and fine-tune 500+ LLMs without writing any code. key features - run models locally on Mac, Windows, Linux - train 500+ models 2x faster with 70% less VRAM - supports GGUF, vision, audio, embedding models - auto-create datasets from PDF, CSV, DOCX - self-healing tool calling and code execution - compare models side by side + export to GGUF to get started, I've shared links in the next tweet.

finetuning LLMs will never be the same! Unsloth just launched an open-source web UI to run and fine-tune 500+ LLMs without writing any code. key features - run models locally on Mac, Windows, Linux - train 500+ models 2x faster with 70% less VRAM - supports GGUF, vision, audio, embedding models - auto-create datasets from PDF, CSV, DOCX - self-healing tool calling and code execution - compare models side by side + export to GGUF to get started, I've shared links in the next tweet.

Akshay 🚀

45,225 views • 3 months ago

Jason Calacanis: Anthropic, OpenAI and others are trying to kill OpenClaw Why? Because an open source agent platform is an existential threat to frontier model companies. @jason: “ People are going to say I'm a conspiracy theorist, but the number one goal, I believe, in the large language model, frontier model space, is to kill (OpenClaw). This is a giant movement to stop it, because this is the equivalent of having an open source Android-like player in the market, and that could be incredibly disruptive. Because, I believe, open source is going to win the day on the large language models and take 90% of the token usage, and I think the entire frontier model space could be undercut by open source. And I think they realize that SLMs, the smaller language models that are verticalized now, that will run on desktops and laptops and are even starting to run on the top ones, that is their biggest competitive threat, and I hope it happens.”

Jason Calacanis: Anthropic, OpenAI and others are trying to kill OpenClaw Why? Because an open source agent platform is an existential threat to frontier model companies. @jason: “ People are going to say I'm a conspiracy theorist, but the number one goal, I believe, in the large language model, frontier model space, is to kill (OpenClaw). This is a giant movement to stop it, because this is the equivalent of having an open source Android-like player in the market, and that could be incredibly disruptive. Because, I believe, open source is going to win the day on the large language models and take 90% of the token usage, and I think the entire frontier model space could be undercut by open source. And I think they realize that SLMs, the smaller language models that are verticalized now, that will run on desktops and laptops and are even starting to run on the top ones, that is their biggest competitive threat, and I hope it happens.”

The All-In Podcast

64,267 views • 3 months ago

Sub-agent Model Selection — Different Tasks, Different Models Your main agent runs Qwen3.6-Plus for quality. But not every subtask needs a flagship model. Now sub-agents can use a different model. Create a skill file with model: openai:qwen3.5-plus and the sub-agent runs on that model. Powerful model for the hard parts, fast model for the easy parts. Save tokens without sacrificing quality on what matters.

Sub-agent Model Selection — Different Tasks, Different Models Your main agent runs Qwen3.6-Plus for quality. But not every subtask needs a flagship model. Now sub-agents can use a different model. Create a skill file with model: openai:qwen3.5-plus and the sub-agent runs on that model. Powerful model for the hard parts, fast model for the easy parts. Save tokens without sacrificing quality on what matters.

Qwen

21,333 views • 3 months ago

You can now run supermemory locally. Introducing the supermemory local - Fully self-contained. Comes with our graph engine, embedding model, etc. - Run on any machine, with your OpenClaw🦞, hermes, claude, etc. - SDKs to add memory to your agent, or build your company brain.

You can now run supermemory locally. Introducing the supermemory local - Fully self-contained. Comes with our graph engine, embedding model, etc. - Run on any machine, with your OpenClaw🦞, hermes, claude, etc. - SDKs to add memory to your agent, or build your company brain.

Dhravya Shah

649,950 views • 1 month ago

Building the Next Evolution of AI in Web3 An AI that automates your daily tasks and professional workflows, no code, no APIs. Owned by its trainer, You. 🫵🏼 Get paid to train the world's first community owned Large Action Model (LAM) Train, Earn, Own. Welcome to Action Model.

Building the Next Evolution of AI in Web3 An AI that automates your daily tasks and professional workflows, no code, no APIs. Owned by its trainer, You. 🫵🏼 Get paid to train the world's first community owned Large Action Model (LAM) Train, Earn, Own. Welcome to Action Model.

Action Model

64,383 views • 11 months ago

Introducing Unsloth Studio ✨ A new open-source web UI to train and run LLMs. • Run models locally on Mac, Windows, Linux • Train 500+ models 2x faster with 70% less VRAM • Supports GGUF, vision, audio, embedding models • Auto-create datasets from PDF, CSV, DOCX • Self-healing tool calling and code execution • Compare models side by side + export to GGUF GitHub: Blog and Guide: Available now on Hugging Face, NVIDIA, Docker and Colab.

Introducing Unsloth Studio ✨ A new open-source web UI to train and run LLMs. • Run models locally on Mac, Windows, Linux • Train 500+ models 2x faster with 70% less VRAM • Supports GGUF, vision, audio, embedding models • Auto-create datasets from PDF, CSV, DOCX • Self-healing tool calling and code execution • Compare models side by side + export to GGUF GitHub: Blog and Guide: Available now on Hugging Face, NVIDIA, Docker and Colab.

Unsloth AI

1,697,662 views • 3 months ago

🚨 KEEP CALM AND AI - AUTONOMOUS AGENTS WITH INFINITE MEMORY We are launching the ability to create arbitrary agents that run on schedule and have access to a persistent and infinite memory The agents will be able to store, retrieve, and update information across sessions and perform repetitive tasks that require persistent memory Another step towards AGI and automating white-collar work 🚀🚀🚀

🚨 KEEP CALM AND AI - AUTONOMOUS AGENTS WITH INFINITE MEMORY We are launching the ability to create arbitrary agents that run on schedule and have access to a persistent and infinite memory The agents will be able to store, retrieve, and update information across sessions and perform repetitive tasks that require persistent memory Another step towards AGI and automating white-collar work 🚀🚀🚀

Bindu Reddy

12,626 views • 5 months ago

🚀 Meta FAIR is releasing several new research artifacts on our road to advanced machine intelligence (AMI). These latest advancements are transforming our understanding of perception. 1️⃣ Meta Perception Encoder: A large-scale vision encoder that excels across several image & video tasks. 2️⃣ Meta Perception Language Model: A fully open & reproducible vision-language model designed to tackle visual recognition tasks. 3️⃣ Meta Locate 3D: An end-to-end model for accurate object localization in 3D environments. 4️⃣ Releasing model weights for our 8B-parameter Dynamic Byte Latent Transformer, an alternative to traditional tokenization methods with the potential to redefine the standards for language model efficiency and reliability. 5️⃣Collaborative Reasoner: A framework for evaluating & improving collaborative reasoning skills in language models. Download the code, datasets, and research papers and learn more about how these artifacts are paving the way for more efficient and accurate AI systems.➡️

🚀 Meta FAIR is releasing several new research artifacts on our road to advanced machine intelligence (AMI). These latest advancements are transforming our understanding of perception. 1️⃣ Meta Perception Encoder: A large-scale vision encoder that excels across several image & video tasks. 2️⃣ Meta Perception Language Model: A fully open & reproducible vision-language model designed to tackle visual recognition tasks. 3️⃣ Meta Locate 3D: An end-to-end model for accurate object localization in 3D environments. 4️⃣ Releasing model weights for our 8B-parameter Dynamic Byte Latent Transformer, an alternative to traditional tokenization methods with the potential to redefine the standards for language model efficiency and reliability. 5️⃣Collaborative Reasoner: A framework for evaluating & improving collaborative reasoning skills in language models. Download the code, datasets, and research papers and learn more about how these artifacts are paving the way for more efficient and accurate AI systems.➡️

AI at Meta

163,214 views • 1 year ago

Devnet is live. Testnet is next. Feature-complete. Fast. Ready to scale. SDKs, APIs, and CLIs already waiting. Start building on Shelby now and get early access to what’s next. More from pranav | Shelby and Marissa Trew 👇

Devnet is live. Testnet is next. Feature-complete. Fast. Ready to scale. SDKs, APIs, and CLIs already waiting. Start building on Shelby now and get early access to what’s next. More from pranav | Shelby and Marissa Trew 👇

Shelby

158,844 views • 6 months ago

Experimenting with custom auto routers for models in Warp. Configure with .yml, and describe what tasks should route to what model in plain english. This example routes frontend tasks to GLM 5.2 and complex architecture decisions to GPT 5.5 high. Wdyt?

Experimenting with custom auto routers for models in Warp. Configure with .yml, and describe what tasks should route to what model in plain english. This example routes frontend tasks to GLM 5.2 and complex architecture decisions to GPT 5.5 high. Wdyt?

Zach Lloyd

11,826 views • 19 days ago

Apple announces Swift Assist for Xcode 16! "Swift Assist serves as a companion for all of a developer’s coding tasks, so they can focus on higher-level problems and solutions. It’s seamlessly integrated into Xcode, and knows the latest software development kits (SDKs) and Swift language features, so developers will always get the latest code features that blend perfectly into their projects. With Swift Assist, tasks like exploring new frameworks and experimenting with new ideas are just one request away. Swift Assist uses a powerful model that runs in the cloud — and like all Apple developer services, it is built with privacy and security in mind. Developers’ code is only used to process requests and never stored on servers, and Apple will not use it to train machine learning models." Swift Assist is coming later this year to Xcode 16.

Apple announces Swift Assist for Xcode 16! "Swift Assist serves as a companion for all of a developer’s coding tasks, so they can focus on higher-level problems and solutions. It’s seamlessly integrated into Xcode, and knows the latest software development kits (SDKs) and Swift language features, so developers will always get the latest code features that blend perfectly into their projects. With Swift Assist, tasks like exploring new frameworks and experimenting with new ideas are just one request away. Swift Assist uses a powerful model that runs in the cloud — and like all Apple developer services, it is built with privacy and security in mind. Developers’ code is only used to process requests and never stored on servers, and Apple will not use it to train machine learning models." Swift Assist is coming later this year to Xcode 16.

M1

79,615 views • 2 years ago

Real-time video analysis with Moondream Point your camera. Ask anything. Detect custom actions. Blazingly fast. Try it →

Real-time video analysis with Moondream Point your camera. Ask anything. Detect custom actions. Blazingly fast. Try it →

moondream

63,309 views • 8 months ago

Nano-Banana moment for Video models soon? In an interview two weeks ago, Demis Hassabis talked about the future of image and video models "When Video converges with Language models you will see some interesting combinaton of Capabilites"

Nano-Banana moment for Video models soon? In an interview two weeks ago, Demis Hassabis talked about the future of image and video models "When Video converges with Language models you will see some interesting combinaton of Capabilites"

AiBattle

47,820 views • 6 months ago

We just released Firefly Custom Models (beta) to everyone 🎉 You can train a model on your own images to keep a consistent style or character! I trained a model on my photography and generated images that would be impossible to get any other way for my upcoming film.

We just released Firefly Custom Models (beta) to everyone 🎉 You can train a model on your own images to keep a consistent style or character! I trained a model on my photography and generated images that would be impossible to get any other way for my upcoming film.

Kris Kashtanova

26,766 views • 3 months ago