Загрузка видео...

Не удалось загрузить видео

На главную

Agentic LLM NPCs with multi-user conversation all running on consumer-grade local hardware

RandoAnalytics's profile picture

stochastics

1,229 subscribers

24,261 просмотров • 1 год назад •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 3

Фото профиля Inno Kean
Inno Kean1 год назад

Why the choice of the Mistral model? Speed?

Фото профиля stochastics
stochastics1 год назад

Mistral 3.1-24B (IQ4 quantized) performed extremely well in my testing for speed, quality (namely with structured responses) and vram usage, given that I’m trying to fit the entire thing into 10gb of vram I’m waiting for @NousResearch to release an even better-performing tuned Hermes@edge variant

Фото профиля Greg Caplan 🚀
Greg Caplan 🚀2 лет назад

Stop wasting time following up with leads. Let our AI agents do it for you.

Похожие видео

Here is Ternary Bonsai 27B running an end-to-end agentic workflow locally with Hermes on an NVIDIA GeForce RTX 5090 GPU. The model reasons, calls tools, reads outputs, modifies files, and surfaces insights - all on consumer hardware, while all private files, intermediate states, and iterations remain local.
1:46

Here is Ternary Bonsai 27B running an end-to-end agentic workflow locally with Hermes on an NVIDIA GeForce RTX 5090 GPU. The model reasons, calls tools, reads outputs, modifies files, and surfaces insights - all on consumer hardware, while all private files, intermediate states, and iterations remain local.

PrismML

151,189 просмотров • 5 дней назад

Introducing `Personalized Agentic RAG` - where the LLM remembers key details about the user and automatically chooses the tool for RAG. We'll build: 👏 ChatGPT like memory ⚙️ Function calling 🧙 Multi-Agent orchestration code:
6:16

Introducing `Personalized Agentic RAG` - where the LLM remembers key details about the user and automatically chooses the tool for RAG. We'll build: 👏 ChatGPT like memory ⚙️ Function calling 🧙 Multi-Agent orchestration code:

Ashpreet Bedi

45,035 просмотров • 2 лет назад

LLama3 + Godot = "living NPCs" 🤯 All running on device. Zero cost. 100% private.
0:54

LLama3 + Godot = "living NPCs" 🤯 All running on device. Zero cost. 100% private.

Danny Postma

331,632 просмотров • 2 лет назад

DeepSeek R1 671B running on 2 M2 Ultras faster than reading speed. Getting close to open-source O1, at home, on consumer hardware. With mlx.distributed and mlx-lm, 3-bit quantization (~4 bpw)
0:54

DeepSeek R1 671B running on 2 M2 Ultras faster than reading speed. Getting close to open-source O1, at home, on consumer hardware. With mlx.distributed and mlx-lm, 3-bit quantization (~4 bpw)

Awni Hannun

866,067 просмотров • 1 год назад

LLM running on Windows 98 PC 26 year old hardware with Intel Pentium II CPU and 128MB RAM. Uses llama98.c, our custom pure C inference engine based on Andrej Karpathy llama2.c Code and DIY guide 👇
0:48

LLM running on Windows 98 PC 26 year old hardware with Intel Pentium II CPU and 128MB RAM. Uses llama98.c, our custom pure C inference engine based on Andrej Karpathy llama2.c Code and DIY guide 👇

EXO Labs

483,118 просмотров • 1 год назад

New course: Build LLM applications that respond to user requests quickly by running on hardware designed for fast inference. This short course was built with Cerebras and taught by Zhenwei Gao, Seb Duerr, and Sarah Chieng. When a model generates text, much of the time is spent moving its weights out of memory and into the compute units. Inference-optimized hardware minimizes that movement, making token generation several times faster than on a typical GPU setup. In this course, the hardware you'll use is Cerebras' Wafer-Scale Engine, which is designed for fast inference by keeping the model's weights close to the compute units. Fast inference makes lengthy agentic workflows go faster, and also unlocks latency-sensitive, real-time applications like live translation and voice agents. Skills you'll gain: - Compare how GPUs, TPUs, and Cerebras' Wafer-Scale Engine each handle the memory-to-compute bottleneck - Build real-time applications powered by fast inference, including personalizing a webpage and running a multi-step workflow to analyze market signals - Adopt concrete habits for agentic coding with fast inference, keeping your sessions focused and steering the model more effectively My teams use Cerebras for several applications that are latency sensitive. Join and build LLM applications that respond quickly:
2:18

New course: Build LLM applications that respond to user requests quickly by running on hardware designed for fast inference. This short course was built with Cerebras and taught by Zhenwei Gao, Seb Duerr, and Sarah Chieng. When a model generates text, much of the time is spent moving its weights out of memory and into the compute units. Inference-optimized hardware minimizes that movement, making token generation several times faster than on a typical GPU setup. In this course, the hardware you'll use is Cerebras' Wafer-Scale Engine, which is designed for fast inference by keeping the model's weights close to the compute units. Fast inference makes lengthy agentic workflows go faster, and also unlocks latency-sensitive, real-time applications like live translation and voice agents. Skills you'll gain: - Compare how GPUs, TPUs, and Cerebras' Wafer-Scale Engine each handle the memory-to-compute bottleneck - Build real-time applications powered by fast inference, including personalizing a webpage and running a multi-step workflow to analyze market signals - Adopt concrete habits for agentic coding with fast inference, keeping your sessions focused and steering the model more effectively My teams use Cerebras for several applications that are latency sensitive. Join and build LLM applications that respond quickly:

Andrew Ng

123,880 просмотров • 2 дней назад

on the frontier of consumer wearables: building an egoless AI product in conversation with Mina Fahmi, building Sandbar between Mina's eloquence and our nyc galavanting, this conversation feels like a work of art. stoked to be able to share it dive in: 00:34 Mina's background 01:42 Robotics, neural interfaces, and hardware 06:06 Bringing nervous system work into designing AI products 10:25 Building with voice 12:38 AI extending self-concept 19:05 AI sycophancy 24:58 Agentic memory 30:11 Hardware decisions at Sandbar 34:16 Hiring at Sandbar 36:04 Bringing AI to more people
40:55

on the frontier of consumer wearables: building an egoless AI product in conversation with Mina Fahmi, building Sandbar between Mina's eloquence and our nyc galavanting, this conversation feels like a work of art. stoked to be able to share it dive in: 00:34 Mina's background 01:42 Robotics, neural interfaces, and hardware 06:06 Bringing nervous system work into designing AI products 10:25 Building with voice 12:38 AI extending self-concept 19:05 AI sycophancy 24:58 Agentic memory 30:11 Hardware decisions at Sandbar 34:16 Hiring at Sandbar 36:04 Bringing AI to more people

ashe

33,719 просмотров • 2 месяцев назад

Build a local LLM app with RAG to chat with Google Drive data using Mistral or Llama-3 running locally on your computer without wiring a single line of Python Code (100% free and without internet):
0:35

Build a local LLM app with RAG to chat with Google Drive data using Mistral or Llama-3 running locally on your computer without wiring a single line of Python Code (100% free and without internet):

Shubham Saboo

189,746 просмотров • 2 лет назад

Yjs is the #1 library for collaborative editing on the web. Today we're announcing built-in Yjs support for Durable Streams. Use it to add multi-user and multi-agent collaboration to your AI apps and agentic systems. Links in the 🧵👇
1:10

Yjs is the #1 library for collaborative editing on the web. Today we're announcing built-in Yjs support for Durable Streams. Use it to add multi-user and multi-agent collaboration to your AI apps and agentic systems. Links in the 🧵👇

Electric

20,177 просмотров • 3 месяцев назад

The best way to learn AI is to build with agents. To help with that, we've launched hands-on labs and a new series on Agentic Engineering. First topic: Agent Skills. Next in the pipeline: planning, context engineering, multi-agent systems, long-running agents,.. Go build!
9:09

The best way to learn AI is to build with agents. To help with that, we've launched hands-on labs and a new series on Agentic Engineering. First topic: Agent Skills. Next in the pipeline: planning, context engineering, multi-agent systems, long-running agents,.. Go build!

elvis

31,802 просмотров • 1 месяц назад

Instant AutoGPT Launcher I wrote a launcher for AutoGPT today. Play with LLM agents using a ComfyUI-like node based interface. Even works with local LLM via ollama (shown in the video) Works on all platforms (windows, linux, mac). Install with 1 click. Use with 0 click.
0:45

Instant AutoGPT Launcher I wrote a launcher for AutoGPT today. Play with LLM agents using a ComfyUI-like node based interface. Even works with local LLM via ollama (shown in the video) Works on all platforms (windows, linux, mac). Install with 1 click. Use with 0 click.

cocktail peanut

11,405 просмотров • 2 лет назад

Preview of a Gaia 🌱 device! Turn-by-turn voice conversation with your own knowledge assistant (a Gaia node). 💸 $15 worth of hardware (pre-tariff!) 🌱 Voice-to-text, LLM and text-to-voice AI models on Gaia nodes 📖 Open source all-the-way — hardware, firmware, server, AI models 🦀 Rust all-the-way from firmware to AI server Notice how I switched to another language (Chinese) in the middle of the conversation and it seamlessly switched too. Very smart. PS. we are still optimizing performance before the software release.
1:24

Preview of a Gaia 🌱 device! Turn-by-turn voice conversation with your own knowledge assistant (a Gaia node). 💸 $15 worth of hardware (pre-tariff!) 🌱 Voice-to-text, LLM and text-to-voice AI models on Gaia nodes 📖 Open source all-the-way — hardware, firmware, server, AI models 🦀 Rust all-the-way from firmware to AI server Notice how I switched to another language (Chinese) in the middle of the conversation and it seamlessly switched too. Very smart. PS. we are still optimizing performance before the software release.

Michael Yuan

14,983 просмотров • 1 год назад

Proud that Microsoft Edge is the first AI browser in the enterprise. - agentic actions - daily briefing - YouTube summarization - multi-tab analysis - search your history with natural language Bringing together the best of AI and enterprise-grade security.
1:00

Proud that Microsoft Edge is the first AI browser in the enterprise. - agentic actions - daily briefing - YouTube summarization - multi-tab analysis - search your history with natural language Bringing together the best of AI and enterprise-grade security.

Mustafa Suleyman

104,008 просмотров • 8 месяцев назад

The top question I got was does Halo support local LLM running on a Mac. Here's an example of the app using gemma-4-26b-a4b-qat via LM Studio running on a base M5/32GB. It's giving an average of 40 tok/s and in my brief testing is able to execute all tool calls correctly. Very usable with this app.
0:37

The top question I got was does Halo support local LLM running on a Mac. Here's an example of the app using gemma-4-26b-a4b-qat via LM Studio running on a base M5/32GB. It's giving an average of 40 tok/s and in my brief testing is able to execute all tool calls correctly. Very usable with this app.

Tanmay

11,351 просмотров • 21 дней назад

Introducing: Cohere Command A+ We’ve created our most powerful LLM yet, optimized it to run on as little hardware as possible, and released it open-source for all.
1:07

Introducing: Cohere Command A+ We’ve created our most powerful LLM yet, optimized it to run on as little hardware as possible, and released it open-source for all.

Cohere

740,455 просмотров • 2 месяцев назад

Introducing Warp 2.0: the Agentic Development Environment 1️⃣ Top overall coding agent: #1 on Terminal-Bench, 71% on SWE-bench Verified 2️⃣ Agent multi-threading: build features, debug, and ship all at once 3️⃣ The first all-in-one platform for agentic development 🧵 Learn more
2:26

Introducing Warp 2.0: the Agentic Development Environment 1️⃣ Top overall coding agent: #1 on Terminal-Bench, 71% on SWE-bench Verified 2️⃣ Agent multi-threading: build features, debug, and ship all at once 3️⃣ The first all-in-one platform for agentic development 🧵 Learn more

Warp

1,445,717 просмотров • 1 год назад

I just deployed an OpenClaw AI agent on NEAR AI running inside TEEs. Real agent. Real memory. Hardware-backed privacy. No prompt leakage. This is what “user-owned AI” actually looks like.
8:49

I just deployed an OpenClaw AI agent on NEAR AI running inside TEEs. Real agent. Real memory. Hardware-backed privacy. No prompt leakage. This is what “user-owned AI” actually looks like.

NEAR DevHub

21,039 просмотров • 5 месяцев назад

Progress on our AI x Hardware project! We have near real-time conversation with both a 2D and 3D animated bot running. These will eventually be full blown paintings you can hang on your wall. Demo videos are unedited. Thoughts below👇
1:10

Progress on our AI x Hardware project! We have near real-time conversation with both a 2D and 3D animated bot running. These will eventually be full blown paintings you can hang on your wall. Demo videos are unedited. Thoughts below👇

Varun Mayya

90,897 просмотров • 2 лет назад

once again causing a scene at the local coffee shop on this week’s delve with mel - 3 words, 3 AI takes: Sean X on whether anthropic should give up on consumer, openai… hardware?, how Apple could do the craziest thing, and the danger of platform lock-in for our memories…
2:20

once again causing a scene at the local coffee shop on this week’s delve with mel - 3 words, 3 AI takes: Sean X on whether anthropic should give up on consumer, openai… hardware?, how Apple could do the craziest thing, and the danger of platform lock-in for our memories…

melody kim

27,044 просмотров • 1 год назад

Top designer at Mercury shows her agentic design explorations in Figma. Designers working on AI features should copy this and sketch how the LLM works before prototyping. 👀 Full conversation on Sneak Peek: carol shows her thought process for designing Mercury Command in Figma👇
4:02

Top designer at Mercury shows her agentic design explorations in Figma. Designers working on AI features should copy this and sketch how the LLM works before prototyping. 👀 Full conversation on Sneak Peek: carol shows her thought process for designing Mercury Command in Figma👇

Jayneil Dalal

19,651 просмотров • 13 дней назад