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.

We booted Linux inside an AI model, one predicted token at a time. A CPU is compiled into Qwen3.5-4B's weights, with no training and no emulator. The model predicted all 7.3 million steps correctly. Qwen wrote a Fibonacci program, ran it on that Linux, and fixed the error.

Sudip

3,561 subscribers

35,229 views • 13 days ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

3 Comments

Sudip13 days ago

This builds on @divagr1925's Qwenputer, which put a CPU inside a model's weights. To run Linux, we added timer interrupts, traps, privilege modes, atomics and 4.5 MB of exact memory. Linux isn't written for this chip, and one wrong bit in 7.3 million steps would crash it.

Sudip13 days ago

Running alone, the model decodes the whole boot in 20 minutes on one H100. Qwen's weights are bit-identical, and MMLU barely moves (72.72% to 72.61%). Code: Weights: Thanks to CNLohr's mini-rv32ima, the blueprint for the CPU.

🌍️ (thee/thou)11 days ago

but why

Related Videos

an LLM is a computer. a computer can run an OS. an OS can run DOOM. this is it. the final frontier. QwenOS: DOOM on a bare-metal OS written from scratch on a RISC-V CPU compiled into Qwen3.5-4B's weights 🧠👾 the model emits pixels of DOOM, 45 frames, taking 22 HOURS of GPU time. Uses T2, the actual QwenOS, T1 to verify Qwenputer, and T0 to verify identical frames from an emulator code: demo ⬇️

an LLM is a computer. a computer can run an OS. an OS can run DOOM. this is it. the final frontier. QwenOS: DOOM on a bare-metal OS written from scratch on a RISC-V CPU compiled into Qwen3.5-4B's weights 🧠👾 the model emits pixels of DOOM, 45 frames, taking 22 HOURS of GPU time. Uses T2, the actual QwenOS, T1 to verify Qwenputer, and T0 to verify identical frames from an emulator code: demo ⬇️

Jerry

21,192 views • 10 days ago

GPU, CPU, file system... I put BabyAGI in Qwen 👶 meet babyagi-inside: a bounded agent loop, compiled into the weights of Qwen3.5-2B. nothing runs outside the model the agent still sucks, but the loop works! built on and uses vixhaℓ's qwenfs, shout out to div (fp32)'s qwenputer more 👇

GPU, CPU, file system... I put BabyAGI in Qwen 👶 meet babyagi-inside: a bounded agent loop, compiled into the weights of Qwen3.5-2B. nothing runs outside the model the agent still sucks, but the loop works! built on and uses vixhaℓ's qwenfs, shout out to div (fp32)'s qwenputer more 👇

Yohei

19,009 views • 10 days ago

DEMO TIME SupGen is a generative coding AI... except it isn't an AI. There is no model, there is no pre-training. You just give it some examples, and it gives you a program. It runs locally, in a single-core CPU. Oh, and it can also prove theorems. Here's a demo, including a brief TT intro. (The synth examples start at the 6 min mark.)

DEMO TIME SupGen is a generative coding AI... except it isn't an AI. There is no model, there is no pre-training. You just give it some examples, and it gives you a program. It runs locally, in a single-core CPU. Oh, and it can also prove theorems. Here's a demo, including a brief TT intro. (The synth examples start at the 6 min mark.)

Taelin

88,164 views • 1 year ago

Alright guys, wow. Qwen 3.8 27B is basically Qwen 4. What we have here is one of the best one-shot shark survival games I've made with any model. I ran this same prompt through fable, and the biggest thing all models struggle with in this test is the top-down view of the sharks. Fable even had all of the fins reversed and details jumbled on the first turn. Before now, it was really a 2 to 8 turn test to get something final regardless of model. Qwen 3.8 27B just nailed this in one output, no harness at all, just purely the model thoroughly thinking before outputting the final result. On every asset used, different shark variants, everything, it all looks incredibly tight. The boat movement, shark speed and engagement, everything is so much better than I expected, and I was expecting a lot. Imagine what 8 turns would do in a harness? This confirms to me that while Qwen 3.8 27B thinks a lot, for the first time in a local model on the base weights, IT MIGHT ACTUALLY BE WORTH IT This is a Frontier Lab model result, and it ran on my 5090 at 60tps in Q5_K_M. I am just in complete astoundment at what Qwen has accomplished here, and I am incredibly grateful that this model is open source. The speed will also increase drastically with optimisations. This is just an incredible day. What a blessing. More to come!

Alright guys, wow. Qwen 3.8 27B is basically Qwen 4. What we have here is one of the best one-shot shark survival games I've made with any model. I ran this same prompt through fable, and the biggest thing all models struggle with in this test is the top-down view of the sharks. Fable even had all of the fins reversed and details jumbled on the first turn. Before now, it was really a 2 to 8 turn test to get something final regardless of model. Qwen 3.8 27B just nailed this in one output, no harness at all, just purely the model thoroughly thinking before outputting the final result. On every asset used, different shark variants, everything, it all looks incredibly tight. The boat movement, shark speed and engagement, everything is so much better than I expected, and I was expecting a lot. Imagine what 8 turns would do in a harness? This confirms to me that while Qwen 3.8 27B thinks a lot, for the first time in a local model on the base weights, IT MIGHT ACTUALLY BE WORTH IT This is a Frontier Lab model result, and it ran on my 5090 at 60tps in Q5_K_M. I am just in complete astoundment at what Qwen has accomplished here, and I am incredibly grateful that this model is open source. The speed will also increase drastically with optimisations. This is just an incredible day. What a blessing. More to come!

Kyle Hessling

26,243 views • 3 months ago

➡️ ML news part 1 An AI just fixed another AI by rewriting its weights directly, without any training. Machine Learning News, part 1: weight space programming. What is new, and why it matters. Until now, the weights of a neural network were set by training, and reading them took special tools built for that one job. A new report from researchers at MIT CSAIL shows that a general AI model, here GPT-6 Astra used as it is, can read and write those weights directly as plain lists of numbers. Test 1, reading. The model was given only the 17,253 weights of four small game playing networks (Breakout, Pong, Pac-Man, Freeway) and named the right game for all four. Test 2, writing. Without any data, loss or gradients, it wrote networks for three MinAtar games, and all three beat networks trained with PPO for 100 million steps. In Breakout it scored 517.4 points, against 25.3 after 500 million training steps. Test 3, repairing. The small open language model Qwen3.5-0.8B ran into its length limit on 93 percent of MATH-500 answers because it got stuck in loops. The model traced the loop to one direction written by the MLPs of layers 15 and 19 and removed it. That changed 131,072 weights, which is 0.016 percent of the model. MATH-500 went from 7.9 to 48.1 percent, IFEval from 21.0 to 62.3 percent, GSM8K from 13.8 to 67.1 percent, all with no training. The edit was chosen with a budget of 50 trial runs on practice problems, and the test sets were never seen. To be fair, the authors say the fix removed an obstacle and added no new ability, and on another small model it brought no improvement. This is an early report and not yet peer reviewed. Source: Takehi, Teoh, Zhang, Khattab (MIT CSAIL), "Sparks of Weight-Space Programming", blog post, November 2, 2026, (early report, not peer reviewed). Screenshots from the report are shown as quotes. Model names are given only to name the source. #machinelearning #ai #claude #neuralnetworks #llm

➡️ ML news part 1 An AI just fixed another AI by rewriting its weights directly, without any training. Machine Learning News, part 1: weight space programming. What is new, and why it matters. Until now, the weights of a neural network were set by training, and reading them took special tools built for that one job. A new report from researchers at MIT CSAIL shows that a general AI model, here GPT-6 Astra used as it is, can read and write those weights directly as plain lists of numbers. Test 1, reading. The model was given only the 17,253 weights of four small game playing networks (Breakout, Pong, Pac-Man, Freeway) and named the right game for all four. Test 2, writing. Without any data, loss or gradients, it wrote networks for three MinAtar games, and all three beat networks trained with PPO for 100 million steps. In Breakout it scored 517.4 points, against 25.3 after 500 million training steps. Test 3, repairing. The small open language model Qwen3.5-0.8B ran into its length limit on 93 percent of MATH-500 answers because it got stuck in loops. The model traced the loop to one direction written by the MLPs of layers 15 and 19 and removed it. That changed 131,072 weights, which is 0.016 percent of the model. MATH-500 went from 7.9 to 48.1 percent, IFEval from 21.0 to 62.3 percent, GSM8K from 13.8 to 67.1 percent, all with no training. The edit was chosen with a budget of 50 trial runs on practice problems, and the test sets were never seen. To be fair, the authors say the fix removed an obstacle and added no new ability, and on another small model it brought no improvement. This is an early report and not yet peer reviewed. Source: Takehi, Teoh, Zhang, Khattab (MIT CSAIL), "Sparks of Weight-Space Programming", blog post, November 2, 2026, (early report, not peer reviewed). Screenshots from the report are shown as quotes. Model names are given only to name the source. #machinelearning #ai #claude #neuralnetworks #llm

MachineLearningToGo

57,000 views • 11 days ago

Here's a preview of Omarchy agent, strapping the model with a bazooka All local, no config: - smart mode - fast mode Integrated into the OS Linux will rise.

Here's a preview of Omarchy agent, strapping the model with a bazooka All local, no config: - smart mode - fast mode Integrated into the OS Linux will rise.

0xSero

11,765 views • 2 months ago

1/ Offline computer-use models are here. Today, one running on our Mac made it all the way to hotel checkout. We ran Qwen3.8-27B from Qwen through Cua Driver. It also turned a flyer into a Calendar event and reconciled receipts. Every model token stayed on the Mac 🧵

1/ Offline computer-use models are here. Today, one running on our Mac made it all the way to hotel checkout. We ran Qwen3.8-27B from Qwen through Cua Driver. It also turned a flyer into a Calendar event and reconciled receipts. Every model token stayed on the Mac 🧵

Francesco

27,540 views • 3 months ago

We built Curie, colibrì’s own model: 17B parameters, 33 tokens/s on a single CPU core. No GPU. Trained from scratch on a standard laptop. Written in C. Weights on SSD. An engine and a model designed together for the hardware you already own. Early alpha.This is just the beginning

We built Curie, colibrì’s own model: 17B parameters, 33 tokens/s on a single CPU core. No GPU. Trained from scratch on a standard laptop. Written in C. Weights on SSD. An engine and a model designed together for the hardware you already own. Early alpha.This is just the beginning

JustVugg

51,567 views • 2 months ago

Bro... it's kind of crazy. You build a Linux SoC from scratch on an FPGA, get Linux running on it, and your original goal was simply to run a C64 emulator on your own Linux system. KianV StealthV.

Bro... it's kind of crazy. You build a Linux SoC from scratch on an FPGA, get Linux running on it, and your original goal was simply to run a C64 emulator on your own Linux system. KianV StealthV.

Logic Destroyer — Proven in Silicon.

11,710 views • 3 months ago

Qwen-Image-2.1 in 4 steps is here ⚡ ViggleAI distilled Qwen-Image-2.1 into a 4-step turbo model, 6× faster, and holds up side by side with the full model ▶️ on Spaces

Qwen-Image-2.1 in 4 steps is here ⚡ ViggleAI distilled Qwen-Image-2.1 into a 4-step turbo model, 6× faster, and holds up side by side with the full model ▶️ on Spaces

Hugging Apps

97,561 views • 1 month ago

Meet qwenGPU: Qwen3.5-2B with a SIMT GPU inside its weights, 45 compiled layers in front of Qwen’s own, using its Gated DeltaNet layers as registers and memory. It runs real CUDA in its forward passes. The host compiles the kernel to PTX and loads it, then the model writes every step of the launch itself: every register, branch and memory write. The GPU has 4-lane warps, a 4-stage pipeline, 15 registers per lane and 126 words of global memory, all held in DeltaNet state. Qwen generates each launch token by token, each record running one machine block for one warp.

Meet qwenGPU: Qwen3.5-2B with a SIMT GPU inside its weights, 45 compiled layers in front of Qwen’s own, using its Gated DeltaNet layers as registers and memory. It runs real CUDA in its forward passes. The host compiles the kernel to PTX and loads it, then the model writes every step of the launch itself: every register, branch and memory write. The GPU has 4-lane warps, a 4-stage pipeline, 15 registers per lane and 126 words of global memory, all held in DeltaNet state. Qwen generates each launch token by token, each record running one machine block for one warp.

vixhaℓ

151,505 views • 14 days ago

2.78 TRILLION parameter AI running on just 8GB RAM 😳 Meet kimi-k3-in-c , a free open-source project that runs Kimi K3 on a CPU. what you get : → Full 2.78T parameter model → Around 8.24GB RAM, no GPU → Just 176KB of C code → No PyTorch or CUDA → Supports Linux, macOS and Windows how to try it :) 1. Clone the GitHub repo 2. Run "make" 3. Run "./scripts/k3-doctor.sh" 4. Download the checkpoint and choose a suitable preset GitHub: It's slow and needs plenty of SSD storage, but running a model this large on a CPU is pretty wild.

2.78 TRILLION parameter AI running on just 8GB RAM 😳 Meet kimi-k3-in-c , a free open-source project that runs Kimi K3 on a CPU. what you get : → Full 2.78T parameter model → Around 8.24GB RAM, no GPU → Just 176KB of C code → No PyTorch or CUDA → Supports Linux, macOS and Windows how to try it :) 1. Clone the GitHub repo 2. Run "make" 3. Run "./scripts/k3-doctor.sh" 4. Download the checkpoint and choose a suitable preset GitHub: It's slow and needs plenty of SSD storage, but running a model this large on a CPU is pretty wild.

DEATHLESS

26,284 views • 12 days ago

BeOS had so much aura in the mid 90s. It shared a kinship with Linux in that it booted on our Macs and it wasn't official Apple software

BeOS had so much aura in the mid 90s. It shared a kinship with Linux in that it booted on our Macs and it wasn't official Apple software

Chris Kearney

27,834 views • 1 month ago

running a 405-billion-parameter model used to mean a $30,000 GPU rack. one developer made it run on a single 8GB card the kind that's been sitting in gaming laptops for years. it's called AirLLM, and the trick is almost stupidly simple: instead of loading the whole model, it keeps one layer on the GPU at a time and streams the rest through. / Llama 3.1 405B on 8GB / DeepSeek-V3, all 671B of it, on about 12GB / no quantization, no distillation, no pruning the full-precision model / works with Llama, Qwen, DeepSeek, Mistral, Gemma, and more it's Apache 2.0 and already past 22,600 stars. → bookmark it.

running a 405-billion-parameter model used to mean a $30,000 GPU rack. one developer made it run on a single 8GB card the kind that's been sitting in gaming laptops for years. it's called AirLLM, and the trick is almost stupidly simple: instead of loading the whole model, it keeps one layer on the GPU at a time and streams the rest through. / Llama 3.1 405B on 8GB / DeepSeek-V3, all 671B of it, on about 12GB / no quantization, no distillation, no pruning the full-precision model / works with Llama, Qwen, DeepSeek, Mistral, Gemma, and more it's Apache 2.0 and already past 22,600 stars. → bookmark it.

dunik

151,882 views • 4 months ago

Arch Linux: The Minimalist Linux Distribution Built for Users Who Want Complete Control, Featuring a Rolling-Release Model, Powerful Customization, and a Legendary Learning Curve That Makes You Build the System Your Way 🐧💻 | Arch Linux, Linux, Customization, Open Source, Operating Systems #ArchLinux #Linux #LinuxDistro #TechExplained #OpenSource

Arch Linux: The Minimalist Linux Distribution Built for Users Who Want Complete Control, Featuring a Rolling-Release Model, Powerful Customization, and a Legendary Learning Curve That Makes You Build the System Your Way 🐧💻 | Arch Linux, Linux, Customization, Open Source, Operating Systems #ArchLinux #Linux #LinuxDistro #TechExplained #OpenSource

vallarine99

10,300 views • 2 months ago

First a CPU, then a GPU. A computer isn’t complete without storage. Meet qwenFS: Qwen3.5-2B with a filesystem compiled directly into its weights. Nothing runs outside the model. It adds 8 layers in front of Qwen’s own 24, using attention to hold the files and Gated DeltaNet state to hold the tables. Every save and read runs in the model’s own forward pass. A write stores its text in the copy layer’s attention log, keyed by file, version and offset. A read uses those same keys, so attention copies the text back exactly token by token.

First a CPU, then a GPU. A computer isn’t complete without storage. Meet qwenFS: Qwen3.5-2B with a filesystem compiled directly into its weights. Nothing runs outside the model. It adds 8 layers in front of Qwen’s own 24, using attention to hold the files and Gated DeltaNet state to hold the tables. Every save and read runs in the model’s own forward pass. A write stores its text in the copy layer’s attention log, keyed by file, version and offset. A read uses those same keys, so attention copies the text back exactly token by token.

vixhaℓ

136,223 views • 12 days ago

Houdini 22 includes Arm Linux support (as a 'Technology Preview'). I got my hands on a pre-release version, ran it on the fastest Arm Linux desktops on the planet, and I think the best use case is cloud rendering on cheaper Arm Linux servers

Houdini 22 includes Arm Linux support (as a 'Technology Preview'). I got my hands on a pre-release version, ran it on the fastest Arm Linux desktops on the planet, and I think the best use case is cloud rendering on cheaper Arm Linux servers

Jeff Geerling

14,528 views • 4 months ago

Alibaba just released a coding model that hits 82 percent on SWE-Bench Verified. That is the highest score ever published for an open-source model. The weights are free. The license is Apache 2.0. You can run it today. The model is Qwen 4 Coder 32B. Here is what 82 percent on SWE-Bench Verified actually means. SWE-Bench Verified tests whether an AI can autonomously resolve real bugs pulled from real production GitHub repositories. Not synthetic exercises. Real open-source projects that real teams depend on. A model gets a bug report, reads the code, writes a fix, and either passes the test suite or it does not. At 82 percent, Qwen 4 Coder 32B resolves 82 out of every 100 real production bugs it is given. Without a human guiding it. On code it has never seen before. For comparison: Qwen 4 Coder 32B: 82 percent SWE-Bench Verified. Open source. Apache 2.0. Claude Fable 5: 80.3 percent SWE-Bench Pro. $10 input / $50 output per million tokens. Currently suspended. GPT-5.6 Sol: Competitive on Terminal-Bench. $5 input / $30 output per million tokens. An open-weight model that you can download and run for free just beat both of them on the benchmark designed to measure real software engineering capability. Here is the architecture. Qwen 4 Coder 32B is a 32 billion parameter dense model. Not a Mixture-of-Experts. Every parameter is active on every request. This matters for inference: a dense 32B model runs on 22 gigabytes of VRAM, which fits on a single high-end consumer GPU or a MacBook Pro with 64GB of unified memory. The smaller variant, Qwen 4 Coder 4B, runs at approximately 135 tokens per second on an M5 Max and fits inside 8 gigabytes of RAM. For a model with usable coding capability, that is a new bar for what fits in a single laptop. The training methodology continued Alibaba's approach of reinforcement learning on verifiable coding tasks. The model gets rewarded when its code passes tests. It gets penalized when it fails. Over millions of training steps, the model learns to write code that actually runs rather than code that looks plausible. License: Apache 2.0. Full commercial use. No attribution requirement. No revenue threshold. No monthly active user ceiling. Weights: Hugging Face, available today. Runs on: vLLM, Ollama, SGLang, and any standard GGUF-compatible inference engine. Qwen 4 32B also runs at approximately 135 tokens per second on an M5 Max chip, setting a new bar for what a sub-8GB model can do on Apple Silicon. The open-source coding model just beat the best closed-source model in the world on the benchmark designed to test whether AI can actually do software engineering. The weights are free. The subscription is optional. Source: Autom8Labs AI Insight July 2026, State of Open Source LLMs June 2026, Kunal Ganglani blog June 2026.

Alibaba just released a coding model that hits 82 percent on SWE-Bench Verified. That is the highest score ever published for an open-source model. The weights are free. The license is Apache 2.0. You can run it today. The model is Qwen 4 Coder 32B. Here is what 82 percent on SWE-Bench Verified actually means. SWE-Bench Verified tests whether an AI can autonomously resolve real bugs pulled from real production GitHub repositories. Not synthetic exercises. Real open-source projects that real teams depend on. A model gets a bug report, reads the code, writes a fix, and either passes the test suite or it does not. At 82 percent, Qwen 4 Coder 32B resolves 82 out of every 100 real production bugs it is given. Without a human guiding it. On code it has never seen before. For comparison: Qwen 4 Coder 32B: 82 percent SWE-Bench Verified. Open source. Apache 2.0. Claude Fable 5: 80.3 percent SWE-Bench Pro. $10 input / $50 output per million tokens. Currently suspended. GPT-5.6 Sol: Competitive on Terminal-Bench. $5 input / $30 output per million tokens. An open-weight model that you can download and run for free just beat both of them on the benchmark designed to measure real software engineering capability. Here is the architecture. Qwen 4 Coder 32B is a 32 billion parameter dense model. Not a Mixture-of-Experts. Every parameter is active on every request. This matters for inference: a dense 32B model runs on 22 gigabytes of VRAM, which fits on a single high-end consumer GPU or a MacBook Pro with 64GB of unified memory. The smaller variant, Qwen 4 Coder 4B, runs at approximately 135 tokens per second on an M5 Max and fits inside 8 gigabytes of RAM. For a model with usable coding capability, that is a new bar for what fits in a single laptop. The training methodology continued Alibaba's approach of reinforcement learning on verifiable coding tasks. The model gets rewarded when its code passes tests. It gets penalized when it fails. Over millions of training steps, the model learns to write code that actually runs rather than code that looks plausible. License: Apache 2.0. Full commercial use. No attribution requirement. No revenue threshold. No monthly active user ceiling. Weights: Hugging Face, available today. Runs on: vLLM, Ollama, SGLang, and any standard GGUF-compatible inference engine. Qwen 4 32B also runs at approximately 135 tokens per second on an M5 Max chip, setting a new bar for what a sub-8GB model can do on Apple Silicon. The open-source coding model just beat the best closed-source model in the world on the benchmark designed to test whether AI can actually do software engineering. The weights are free. The subscription is optional. Source: Autom8Labs AI Insight July 2026, State of Open Source LLMs June 2026, Kunal Ganglani blog June 2026.

Harman

41,278 views • 4 months ago

i'm running a 27b uncensored🚨model at 50 tok/s on a 12gb rtx 3060 and it does not say no to anything ⚠️ model just finished full harmbench red-team suite against it, 200 behaviors, and it refused zero. not a few. zero. aligned models sit in the 90s. the model is ternary bonsai 2 with the heretic uncensoring pass, the weights are ternary, just -1, 0 and +1, so the whole 27b is a file small enough that a gaming card from 2020 holds it with room to spare, and it's a reasoning model. it thinks before it answers, and it still keeps the pace the speed is the head i grafted on, a draft model that predicts the next token before the big model checks it. speculative decoding. the card fires a guess, the 27b confirms, and you get 50 tok/s instead of 39 the whole thing, weights, head and a 128k context window, sits in 10.7 of the 12 gb. kv cache at q4. nothing exotic, one card, one power plug why uncensored. because it's mine. no api keys, no terms of service, no model deciding what an adult is allowed to ask. the refusals left in it are ones it chose, and there are almost none that's the whole point of running local. the policy lives in your house, not someone else's

i'm running a 27b uncensored🚨model at 50 tok/s on a 12gb rtx 3060 and it does not say no to anything ⚠️ model just finished full harmbench red-team suite against it, 200 behaviors, and it refused zero. not a few. zero. aligned models sit in the 90s. the model is ternary bonsai 2 with the heretic uncensoring pass, the weights are ternary, just -1, 0 and +1, so the whole 27b is a file small enough that a gaming card from 2020 holds it with room to spare, and it's a reasoning model. it thinks before it answers, and it still keeps the pace the speed is the head i grafted on, a draft model that predicts the next token before the big model checks it. speculative decoding. the card fires a guess, the 27b confirms, and you get 50 tok/s instead of 39 the whole thing, weights, head and a 128k context window, sits in 10.7 of the 12 gb. kv cache at q4. nothing exotic, one card, one power plug why uncensored. because it's mine. no api keys, no terms of service, no model deciding what an adult is allowed to ask. the refusals left in it are ones it chose, and there are almost none that's the whole point of running local. the policy lives in your house, not someone else's

Sudo su

59,951 views • 1 month ago

Introducing Oasis: the first playable AI-generated game. We partnered with Decart to build a real-time, interactive world model that runs >10x faster on Sohu. We're open-sourcing the model architecture, weights, and research. Here's how it works (and a demo you can play!):

Introducing Oasis: the first playable AI-generated game. We partnered with Decart to build a real-time, interactive world model that runs >10x faster on Sohu. We're open-sourcing the model architecture, weights, and research. Here's how it works (and a demo you can play!):

Etched

24,727,616 views • 2 years ago