Загрузка видео...

Не удалось загрузить видео

На главную

Auto regressive LLMs are officially on notice. run Gemma 4 26B diffusion gguf with llama.cpp Google just dropped DiffusionGemma-26B, and it completely flips how we generate text. instead of predicting words one by one, it generates 256 tokens in parallel using bi-directional attention. its like stable diffusion, but for...

analogalok's profile picture

Alok

1,226 subscribers

52,656 просмотров • 1 месяц назад •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

six months ago this wasn't happening on 8gb vram. running unsloth's Q4_K_XL quant of gemma 4 26b-a4b-it-qat, a sparse MoE model with only 4b active params on a single rtx 4060 laptop gpu, 8gb vram, 20+ tok/s decode. no cloud, no api, no offload hacks. just a gaming laptop on battery. what makes it fit: google's QAT (quantization aware training), plus MTP (multi token prediction) support in the latest llama.cpp builds. that combo is the single biggest unlock for local inference on low vram. rtx 3060, rtx 3070, gtx 1070, gtx 1080, rtx 4050, rtx 4060, rtx 5050, rtx 5060 — any 6-8gb consumer gpu, old or new — this model runs on it. world cup season, so i told it to build a soccer themed flappy bird clone. one shot, zero iteration, fully playable. six months ago an 8gb model could barely clone vanilla flappy bird. now it's shipping a themed game from a sparse MoE model running locally on a laptop battery. inference benchmarks: - decode throughput: 30 tok/s - context: 64k. this is the real unlock. 64k ctx is what makes a hermes agent loop viable locally on this model, not just single-turn chat. llama.cpp flags: -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf -c 64000 -cmoe --port 8080 game's deployed on my own site, built and shipped end to end with open source llm, zero closed source api dependency in the pipeline. link in the description. gguf weights on huggingface, link in the comments. pull it down, run it on whatever 8gb card is sitting in your rig. try the game and tell me your score and what you want in v2. local llms on consumer gpus stopped being a meme.
0:20

six months ago this wasn't happening on 8gb vram. running unsloth's Q4_K_XL quant of gemma 4 26b-a4b-it-qat, a sparse MoE model with only 4b active params on a single rtx 4060 laptop gpu, 8gb vram, 20+ tok/s decode. no cloud, no api, no offload hacks. just a gaming laptop on battery. what makes it fit: google's QAT (quantization aware training), plus MTP (multi token prediction) support in the latest llama.cpp builds. that combo is the single biggest unlock for local inference on low vram. rtx 3060, rtx 3070, gtx 1070, gtx 1080, rtx 4050, rtx 4060, rtx 5050, rtx 5060 — any 6-8gb consumer gpu, old or new — this model runs on it. world cup season, so i told it to build a soccer themed flappy bird clone. one shot, zero iteration, fully playable. six months ago an 8gb model could barely clone vanilla flappy bird. now it's shipping a themed game from a sparse MoE model running locally on a laptop battery. inference benchmarks: - decode throughput: 30 tok/s - context: 64k. this is the real unlock. 64k ctx is what makes a hermes agent loop viable locally on this model, not just single-turn chat. llama.cpp flags: -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf -c 64000 -cmoe --port 8080 game's deployed on my own site, built and shipped end to end with open source llm, zero closed source api dependency in the pipeline. link in the description. gguf weights on huggingface, link in the comments. pull it down, run it on whatever 8gb card is sitting in your rig. try the game and tell me your score and what you want in v2. local llms on consumer gpus stopped being a meme.

Alok

60,725 просмотров • 23 дней назад

HERMES AGENT NOW RUNS ON AN 8GB LAPTOP GPU JUST AS EASILY AS IT RUNS ON A 128GB MINI PC Nous Research shipped the official Hermes Agent Desktop App this week. Someone pointed it at a local llama server running on an RTX 4060 with 16GB system RAM. The integration took two minutes The model behind it: Gemma 4 26B MoE, QAT quantized, running on 8GB of VRAM. A 60k token prompt held a stable 20 tokens a second, flat, no slowdown as context grew. The flags were nothing exotic, just -cmoe -c 248000 on llama.cpp What that 8GB setup does out of the box: reads and patches its own code, runs it in a terminal, debugs errors, manages GitHub repos, spawns sub-agents for parallel work. Browses the web with vision to debug a UI. Schedules cron jobs in plain language. Connects to Notion, Google Workspace, Linear, and Obsidian to manage tasks on its own That's the same agent layer running on a Minisforum MS-S1 MAX with 128GB of unified memory, 96GB of it to the GPU, holding a 120B model at 56 tokens a second instead of a 26B model at 20. Same software, same tool execution, same zero API key. The only thing that changes between an $800 laptop and a $2,000 mini PC is how big a model you can afford to run underneath it The barrier to running a real autonomous agent locally didn't just drop. It dropped all the way down to hardware most people already own
0:20

HERMES AGENT NOW RUNS ON AN 8GB LAPTOP GPU JUST AS EASILY AS IT RUNS ON A 128GB MINI PC Nous Research shipped the official Hermes Agent Desktop App this week. Someone pointed it at a local llama server running on an RTX 4060 with 16GB system RAM. The integration took two minutes The model behind it: Gemma 4 26B MoE, QAT quantized, running on 8GB of VRAM. A 60k token prompt held a stable 20 tokens a second, flat, no slowdown as context grew. The flags were nothing exotic, just -cmoe -c 248000 on llama.cpp What that 8GB setup does out of the box: reads and patches its own code, runs it in a terminal, debugs errors, manages GitHub repos, spawns sub-agents for parallel work. Browses the web with vision to debug a UI. Schedules cron jobs in plain language. Connects to Notion, Google Workspace, Linear, and Obsidian to manage tasks on its own That's the same agent layer running on a Minisforum MS-S1 MAX with 128GB of unified memory, 96GB of it to the GPU, holding a 120B model at 56 tokens a second instead of a 26B model at 20. Same software, same tool execution, same zero API key. The only thing that changes between an $800 laptop and a $2,000 mini PC is how big a model you can afford to run underneath it The barrier to running a real autonomous agent locally didn't just drop. It dropped all the way down to hardware most people already own

NO1ennn

40,079 просмотров • 24 дней назад

New episode of the Information Bottleneck! We talked with Stefano Ermon about why he thinks diffusion LLMs will replace autoregressive ones. Stefano co-invented DDIM, FlashAttention, DPO, and score-based diffusion models. He's a Stanford professor and now runs Inception AI, where they built Mercury II. We go deep but also cover the bigger picture - the startup journey, PhD vs industry, and where AI is heading. A few things that stuck with me: - He thinks of autoregressive models as typewriters and diffusion models as editors. One goes left to right. The other starts messy and refines. - Mercury II (their text difussion model) already beats the fastest autoregressive models on latency-critical stuff as voice agents, code suggestions, anything where you have a tight time budget. And it does it because diffusion generates tokens in parallel instead of one at a time. - We also got into whether AI will actually replace software engineers (his answer: no), PhD vs industry advice, and what it was like going from an ICML best paper to raising money.
0:54

New episode of the Information Bottleneck! We talked with Stefano Ermon about why he thinks diffusion LLMs will replace autoregressive ones. Stefano co-invented DDIM, FlashAttention, DPO, and score-based diffusion models. He's a Stanford professor and now runs Inception AI, where they built Mercury II. We go deep but also cover the bigger picture - the startup journey, PhD vs industry, and where AI is heading. A few things that stuck with me: - He thinks of autoregressive models as typewriters and diffusion models as editors. One goes left to right. The other starts messy and refines. - Mercury II (their text difussion model) already beats the fastest autoregressive models on latency-critical stuff as voice agents, code suggestions, anything where you have a tight time budget. And it does it because diffusion generates tokens in parallel instead of one at a time. - We also got into whether AI will actually replace software engineers (his answer: no), PhD vs industry advice, and what it was like going from an ICML best paper to raising money.

Ravid Shwartz Ziv

22,350 просмотров • 3 месяцев назад

a new 8GB VRAM GPU dense Local LLM leader was born yesterday runs on: RTX 4060 / RTX 3070 / RTX 2080. any 8GB card Qwen 3.5 9B (dense) was the go to for 6-8GB VRAM builds. Gemma 4 12B QAT (dense) just changed that. same llama.cpp + cuda 13.2. i7 12700H. 16GB RAM. same -ngl 99 flags. same 48k context. unsloth gemma-4-12b-it-Q4_K_M.gguf → 15 tok/sec @ 48k ctx unsloth gemma-4-12B-it-qat-UD-Q4_K_XL.gguf → 32 tok/sec @ 48k ctx → 26 tok/sec @ 64k ctx 64k context is a big deal. Hermes 3 agent requires 64k minimum to run. you're now getting full hermes compatible context on a budget consumer GPU at 26 tok/sec locally. 2.1x faster on identical hardware. and here's the part that breaks your brain: the QAT-UD-Q4_K_XL is actually SMALLER than the Q4_K_M "XL" why? QAT = Quantization Aware Training Google didn't train the model first and compress it later they trained it to be quantized from day one the weights already know how to survive low precision that's why you get more quality per byte llamacpp flags: -m gemma-4-12B-it-qat-UD-Q4_K_XL.gguf -cnv -ngl 99 -c 48000 -v fits in 8GB VRAM clean. no API. no cloud. no subscription. and this isn't even the MTP variant yet Gemma-4-E2B QAT runs on 3GB RAM, E4B on 5GB, 12B on 7GB, 26-A4B on 15GB and 31B on 18GB. I have benchmarked the 26b and 31b qat as well on a single RTX 4090, checkout the comments for details. If you have a 6GB or 8GB VRAM GPU, post your numbers. more benchmarks and configs coming soon
0:22

a new 8GB VRAM GPU dense Local LLM leader was born yesterday runs on: RTX 4060 / RTX 3070 / RTX 2080. any 8GB card Qwen 3.5 9B (dense) was the go to for 6-8GB VRAM builds. Gemma 4 12B QAT (dense) just changed that. same llama.cpp + cuda 13.2. i7 12700H. 16GB RAM. same -ngl 99 flags. same 48k context. unsloth gemma-4-12b-it-Q4_K_M.gguf → 15 tok/sec @ 48k ctx unsloth gemma-4-12B-it-qat-UD-Q4_K_XL.gguf → 32 tok/sec @ 48k ctx → 26 tok/sec @ 64k ctx 64k context is a big deal. Hermes 3 agent requires 64k minimum to run. you're now getting full hermes compatible context on a budget consumer GPU at 26 tok/sec locally. 2.1x faster on identical hardware. and here's the part that breaks your brain: the QAT-UD-Q4_K_XL is actually SMALLER than the Q4_K_M "XL" why? QAT = Quantization Aware Training Google didn't train the model first and compress it later they trained it to be quantized from day one the weights already know how to survive low precision that's why you get more quality per byte llamacpp flags: -m gemma-4-12B-it-qat-UD-Q4_K_XL.gguf -cnv -ngl 99 -c 48000 -v fits in 8GB VRAM clean. no API. no cloud. no subscription. and this isn't even the MTP variant yet Gemma-4-E2B QAT runs on 3GB RAM, E4B on 5GB, 12B on 7GB, 26-A4B on 15GB and 31B on 18GB. I have benchmarked the 26b and 31b qat as well on a single RTX 4090, checkout the comments for details. If you have a 6GB or 8GB VRAM GPU, post your numbers. more benchmarks and configs coming soon

Alok

259,993 просмотров • 1 месяц назад

karpathy just broke the internet with something called auto research it’s basically an ai research agent that runs experiments for you 24/7 you give it a goal like “make this model better” “find a higher converting landing page” “lower customer acquisition cost” then it runs a loop: 1) plan an experiment 2) edit the code or config 3) run a short test on a gpu 4) read the metrics 5) keep the winner 6) try again over and over while you sleep by the morning you wake up to the best version actual tested improvements think of it like a robot research intern that runs hundreds of experiments and only keeps the winners this is link to his repo for your to mess around with it in the latest episode of The Startup Ideas Podcast (SIP) 🧃 i break down: • what auto research actually is • how it works step by step • 10 business ideas you can build with it • how to install it and start using it this one is saucy because tools like this change how startups get built watch
24:21

karpathy just broke the internet with something called auto research it’s basically an ai research agent that runs experiments for you 24/7 you give it a goal like “make this model better” “find a higher converting landing page” “lower customer acquisition cost” then it runs a loop: 1) plan an experiment 2) edit the code or config 3) run a short test on a gpu 4) read the metrics 5) keep the winner 6) try again over and over while you sleep by the morning you wake up to the best version actual tested improvements think of it like a robot research intern that runs hundreds of experiments and only keeps the winners this is link to his repo for your to mess around with it in the latest episode of The Startup Ideas Podcast (SIP) 🧃 i break down: • what auto research actually is • how it works step by step • 10 business ideas you can build with it • how to install it and start using it this one is saucy because tools like this change how startups get built watch

GREG ISENBERG

435,542 просмотров • 4 месяцев назад

i watched gemma 4 12b build something genuinely impressive today, and then loop itself to death right in front of me. the full run is in the video, sped up but completely uncut, watch it to the end and you will catch the exact moment it stops building and starts looping right in the middle of the work. the task was clean, build a single file gravity simulator, n-body physics, orbits, collisions, running locally on one 3090 through an agent. and for ten minutes it was a joy to watch. it reached for a symplectic integrator on its own, the correct one, the kind that keeps orbits stable instead of spiralling out. real gravity with softening, proper orbital velocities, momentum conserved on collision. the physics was right. the thing actually worked. then on the very last step, writing a few tests to prove its own code, it fell into a loop. not a crash, a loop. it started repeating itself and would not stop. ten more minutes, thirty four thousand tokens into a single answer, the same fragments over and over, until i killed it myself. so it's not that gemma can't code. it did the hard part beautifully. it cannot finish. it cannot hold a long task together without unravelling, and finishing is the entire job in agentic work. here's the part that stings. i run this exact task, same harness, same card, on the chinese open models, qwen especially, and i never see this. they build it, they test it, they stop. every single time. google has the raw capability, you can see it sitting right there in the code, and then the model loops itself to death on a task a 27b from alibaba finishes clean. open weights, apache 2.0, so much to love on paper. i just need it to know when to stop talking.
4:15

i watched gemma 4 12b build something genuinely impressive today, and then loop itself to death right in front of me. the full run is in the video, sped up but completely uncut, watch it to the end and you will catch the exact moment it stops building and starts looping right in the middle of the work. the task was clean, build a single file gravity simulator, n-body physics, orbits, collisions, running locally on one 3090 through an agent. and for ten minutes it was a joy to watch. it reached for a symplectic integrator on its own, the correct one, the kind that keeps orbits stable instead of spiralling out. real gravity with softening, proper orbital velocities, momentum conserved on collision. the physics was right. the thing actually worked. then on the very last step, writing a few tests to prove its own code, it fell into a loop. not a crash, a loop. it started repeating itself and would not stop. ten more minutes, thirty four thousand tokens into a single answer, the same fragments over and over, until i killed it myself. so it's not that gemma can't code. it did the hard part beautifully. it cannot finish. it cannot hold a long task together without unravelling, and finishing is the entire job in agentic work. here's the part that stings. i run this exact task, same harness, same card, on the chinese open models, qwen especially, and i never see this. they build it, they test it, they stop. every single time. google has the raw capability, you can see it sitting right there in the code, and then the model loops itself to death on a task a 27b from alibaba finishes clean. open weights, apache 2.0, so much to love on paper. i just need it to know when to stop talking.

Sudo su

39,574 просмотров • 1 месяц назад

QVAC SDK 0.14.0 is live. This release makes the on-device stack faster on mobile, ships the developer-agent path, and takes local text-to-speech to 31 languages. Main highlights: - OpenCode and OpenClaw. The first official OpenCode plugin, plus a maintained OpenClaw compatibility path, both built on managed mode and qvac serve. Point a coding agent at a local model with far less setup and far fewer surprises. - Brain-computer interface transcription, on the SDK. Take recorded neural signal data and decode it into text, fully on-device, no cloud. Stream it in chunks through a simple API. In 0.14 it runs GPU-accelerated on iOS. - Text to Speech in 31 languages with our Supertonic3 upgrade. VOICE AND SPEECH - Supertonic3 multilingual TTS, 5 languages to 31. - Chatterbox and Supertonic now run on the Android GPU, with lower memory use (especially on iOS), quantized s3gen Chatterbox support, and a fix for Chatterbox occasionally emitting random speech. - Whisper transcription now runs on the iOS GPU. Parakeet runs on the Android GPU, with steadier real-time streaming. VISION AND OCR - VLM multi-tile batching: high-resolution Pan and Scan images are encoded in one pass instead of tile by tile, for faster vision throughput. - OCR on ggml (EasyOCR and DocTR) reaches full speed parity with the onnx path, across Metal, OpenCL, and Vulkan. PLATFORM AND RELIABILITY - Dynamic compute backends on Linux: one build picks the right backend at runtime, and opens the door to ROCm and CUDA support without per-backend builds. - Thinking tokens are kept out of the model context, so reasoning no longer fills the KV cache. SDK 0.14.0 is now leaner and faster to start. Let’s build.
1:51

QVAC SDK 0.14.0 is live. This release makes the on-device stack faster on mobile, ships the developer-agent path, and takes local text-to-speech to 31 languages. Main highlights: - OpenCode and OpenClaw. The first official OpenCode plugin, plus a maintained OpenClaw compatibility path, both built on managed mode and qvac serve. Point a coding agent at a local model with far less setup and far fewer surprises. - Brain-computer interface transcription, on the SDK. Take recorded neural signal data and decode it into text, fully on-device, no cloud. Stream it in chunks through a simple API. In 0.14 it runs GPU-accelerated on iOS. - Text to Speech in 31 languages with our Supertonic3 upgrade. VOICE AND SPEECH - Supertonic3 multilingual TTS, 5 languages to 31. - Chatterbox and Supertonic now run on the Android GPU, with lower memory use (especially on iOS), quantized s3gen Chatterbox support, and a fix for Chatterbox occasionally emitting random speech. - Whisper transcription now runs on the iOS GPU. Parakeet runs on the Android GPU, with steadier real-time streaming. VISION AND OCR - VLM multi-tile batching: high-resolution Pan and Scan images are encoded in one pass instead of tile by tile, for faster vision throughput. - OCR on ggml (EasyOCR and DocTR) reaches full speed parity with the onnx path, across Metal, OpenCL, and Vulkan. PLATFORM AND RELIABILITY - Dynamic compute backends on Linux: one build picks the right backend at runtime, and opens the door to ROCm and CUDA support without per-backend builds. - Thinking tokens are kept out of the model context, so reasoning no longer fills the KV cache. SDK 0.14.0 is now leaner and faster to start. Let’s build.

QVAC

23,973,950 просмотров • 15 дней назад

SORA by Hand ✍️ OpenAI’s #SORA took over the Internet when it was announced earlier this year. The technology behind Sora is the Diffusion Transformer (DiT) developed by William Peebles and Shining Xie. How does DiT work? 𝗚𝗼𝗮𝗹: Generate a video conditioned by a text prompt and a series of diffusion steps [1] Given ↳ Video ↳ Prompt: "sora is sky" ↳ Diffusion step: t = 3 [2] Video → Patches ↳ Divide all pixels in all frames into 4 spacetime patches [3] Visual Encoder: Pixels 🟨 → Latent 🟩 ↳ Multiply the patches with weights and biases, followed by ReLU ↳ The result is a latent feature vector per patch ↳ The purpose is dimension reduction from 4 (2x2x1) to 2 (2x1). ↳ In the paper, the reduction is 196,608 (256x256x3)→ 4096 (32x32x4) [4] ⬛ Add Noise ↳ Sample a noise according to the diffusion time step t. Typically, the larger the t, the smaller the noise. ↳ Add the Sampled Noise to latent features to obtain Noised Latent. ↳ The goal is to purposely add noise to a video and ask the model to guess what that noise is. ↳ This is analogous to training a language model by purposely deleting a word in a sentence and ask the model to guess what the deleted word was. [5-7] 🟪 Conditioning by Adaptive Layer Norm [5] Encode Conditions ↳ Encode "sora is sky" into a text embedding vector [0,1,-1]. ↳ Encode t = 3 to as a binary vector [1,1]. ↳ Concatenate the two vectors in to a 5D column vector. [6] Estimate Scale/Shift ↳ Multiply the combined vector with weights and biases ↳ The goal is to estimate the scale [2,-1] and shift [-1,5]. ↳ Copy the result to (X) and (+) [7] Apply Scale/Sift ↳ Scale the noised latent by [2,-1] ↳ Shifted the scaled noised latent by [-1, 5] ↳ The result is "conditioned" noise latent. [8-10] Transformer [8] Self-Attention ↳ Feed the conditioned noised latent to Query-Key function to obtain a self-attention matrix ↳ Value is omitted for simplicity [9] Attention Pooling ↳ Multiply the conditioned noised latent with the self-attention matrix ↳ The result are attention weighted features [10] Pointwise Feed Forward Network ↳ Multiply the attention weighted features with weights and biases ↳ The result is the Predicted Noise 🏋️‍♂️ 𝗧𝗿𝗮𝗶𝗻 [11] ↳ Calculate MSE loss gradients by taking the different between the Predicted Noise and the Sampled Noise (ground truth). ↳ Use the loss gradients to kick off backpropagation to update all learnable parameters (red borders) ↳ Note the visual encoder and decoder's parameters are frozen (blue borders) 🎨 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗲 (𝗦𝗮𝗺𝗽𝗹𝗲) [12] Denoise ↳ Subtract the predicted noise from the noised latent to obtain the noise-free latent [13] Visual Decoder: Latent 🟩 → Pixels 🟨 ↳ Multiply the patches with weights and biases, followed by ReLU [14] Patches → Video ↳ Rearrange patches into a sequence of video frames.
0:43

SORA by Hand ✍️ OpenAI’s #SORA took over the Internet when it was announced earlier this year. The technology behind Sora is the Diffusion Transformer (DiT) developed by William Peebles and Shining Xie. How does DiT work? 𝗚𝗼𝗮𝗹: Generate a video conditioned by a text prompt and a series of diffusion steps [1] Given ↳ Video ↳ Prompt: "sora is sky" ↳ Diffusion step: t = 3 [2] Video → Patches ↳ Divide all pixels in all frames into 4 spacetime patches [3] Visual Encoder: Pixels 🟨 → Latent 🟩 ↳ Multiply the patches with weights and biases, followed by ReLU ↳ The result is a latent feature vector per patch ↳ The purpose is dimension reduction from 4 (2x2x1) to 2 (2x1). ↳ In the paper, the reduction is 196,608 (256x256x3)→ 4096 (32x32x4) [4] ⬛ Add Noise ↳ Sample a noise according to the diffusion time step t. Typically, the larger the t, the smaller the noise. ↳ Add the Sampled Noise to latent features to obtain Noised Latent. ↳ The goal is to purposely add noise to a video and ask the model to guess what that noise is. ↳ This is analogous to training a language model by purposely deleting a word in a sentence and ask the model to guess what the deleted word was. [5-7] 🟪 Conditioning by Adaptive Layer Norm [5] Encode Conditions ↳ Encode "sora is sky" into a text embedding vector [0,1,-1]. ↳ Encode t = 3 to as a binary vector [1,1]. ↳ Concatenate the two vectors in to a 5D column vector. [6] Estimate Scale/Shift ↳ Multiply the combined vector with weights and biases ↳ The goal is to estimate the scale [2,-1] and shift [-1,5]. ↳ Copy the result to (X) and (+) [7] Apply Scale/Sift ↳ Scale the noised latent by [2,-1] ↳ Shifted the scaled noised latent by [-1, 5] ↳ The result is "conditioned" noise latent. [8-10] Transformer [8] Self-Attention ↳ Feed the conditioned noised latent to Query-Key function to obtain a self-attention matrix ↳ Value is omitted for simplicity [9] Attention Pooling ↳ Multiply the conditioned noised latent with the self-attention matrix ↳ The result are attention weighted features [10] Pointwise Feed Forward Network ↳ Multiply the attention weighted features with weights and biases ↳ The result is the Predicted Noise 🏋️‍♂️ 𝗧𝗿𝗮𝗶𝗻 [11] ↳ Calculate MSE loss gradients by taking the different between the Predicted Noise and the Sampled Noise (ground truth). ↳ Use the loss gradients to kick off backpropagation to update all learnable parameters (red borders) ↳ Note the visual encoder and decoder's parameters are frozen (blue borders) 🎨 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗲 (𝗦𝗮𝗺𝗽𝗹𝗲) [12] Denoise ↳ Subtract the predicted noise from the noised latent to obtain the noise-free latent [13] Visual Decoder: Latent 🟩 → Pixels 🟨 ↳ Multiply the patches with weights and biases, followed by ReLU [14] Patches → Video ↳ Rearrange patches into a sequence of video frames.

Tom Yeh

238,166 просмотров • 2 лет назад

If you could only learn one thing that will be relevant for the next 10-20 years, focus on learning how to deal with data. The future is not about faster hardware, smarter algorithms, or better ideas. The future is about DATA, and those who know how to deal with it will stay relevant much longer than anyone else. I recorded a video to show you how easy it is to get started. In the video, I'm using Kestra. For a long time, I was a fan of AirFlow. Then, I moved to AWS Step Functions. Today, I only use Kestra. Kestra is open-source (repo link below) and kind enough to sponsor my work. The video will show you how easy it is to do the following: 1. Run Kestra locally (literally, one command) 2. Build a simple flow 3. Run Python scripts as part of your flow 4. Connect to HuggingFace models If you have never built a data pipeline, open Kestra's Quick Start Guide and follow their examples. (I think it will take you one weekend to feel comfortable with the application and build the courage you need to get into more serious work.)
5:18

If you could only learn one thing that will be relevant for the next 10-20 years, focus on learning how to deal with data. The future is not about faster hardware, smarter algorithms, or better ideas. The future is about DATA, and those who know how to deal with it will stay relevant much longer than anyone else. I recorded a video to show you how easy it is to get started. In the video, I'm using Kestra. For a long time, I was a fan of AirFlow. Then, I moved to AWS Step Functions. Today, I only use Kestra. Kestra is open-source (repo link below) and kind enough to sponsor my work. The video will show you how easy it is to do the following: 1. Run Kestra locally (literally, one command) 2. Build a simple flow 3. Run Python scripts as part of your flow 4. Connect to HuggingFace models If you have never built a data pipeline, open Kestra's Quick Start Guide and follow their examples. (I think it will take you one weekend to feel comfortable with the application and build the courage you need to get into more serious work.)

Santiago

51,012 просмотров • 1 год назад

We've officially released and open-sourced HunyuanImage 2.1, our latest text-to-image model. The new model delivers on our commitment to balancing performance and quality. With native 2K image generation, HunyuanImage 2.1 is an advanced open-source text-to-image model.🎨 ✨ New in 2.1: 🔹Advanced Semantics: Supports ultra-long and complex prompts of up to 1000 tokens, and precisely controls the generation of multiple subjects in a single image. 🔹Precise Chinese and English Text Rendering with seamless image–text integration: The model naturally integrates text into images, making it suitable for a wide range of applications such as product covers, illustrations, and poster design to meet the needs of various fields. 🔹Rich Styles and High Aesthetic: Capable of generating images in various styles—including photorealistic portraits, comics, and vinyl figures—it delivers outstanding visual appeal and artistic quality. 🔹High-Quality Generation: Efficiently produces ultra-high-definition (2K) images in the same time other models take to generate a 1K image. HunyuanImage 2.1 uses two text encoders: a multimodal large language model (MLLM) to improve the model's image and text alignment capabilities, and a multi-language character-aware encoder to improve text rendering capabilities. The model is a single- and double-stream diffusion transformer with 17B parameters. We've also open-sourced the weights of the the accelerated version with meanflow which reduces inference steps from 100 to just 8, and PromptEnhancer, the first industrial-grade rewriting model that enhances your prompts for more nuanced and expressive image generation. Now, creators turn complex ideas—like posters with slogans or multi-panel comics—into visuals faster than ever. We’re just getting started. Stay tuned for our native multimodal image generation model coming soon. 🌐Website: 🔗Github: 🤗Hugging Face: ✨Hugging Face Demo:
1:39

We've officially released and open-sourced HunyuanImage 2.1, our latest text-to-image model. The new model delivers on our commitment to balancing performance and quality. With native 2K image generation, HunyuanImage 2.1 is an advanced open-source text-to-image model.🎨 ✨ New in 2.1: 🔹Advanced Semantics: Supports ultra-long and complex prompts of up to 1000 tokens, and precisely controls the generation of multiple subjects in a single image. 🔹Precise Chinese and English Text Rendering with seamless image–text integration: The model naturally integrates text into images, making it suitable for a wide range of applications such as product covers, illustrations, and poster design to meet the needs of various fields. 🔹Rich Styles and High Aesthetic: Capable of generating images in various styles—including photorealistic portraits, comics, and vinyl figures—it delivers outstanding visual appeal and artistic quality. 🔹High-Quality Generation: Efficiently produces ultra-high-definition (2K) images in the same time other models take to generate a 1K image. HunyuanImage 2.1 uses two text encoders: a multimodal large language model (MLLM) to improve the model's image and text alignment capabilities, and a multi-language character-aware encoder to improve text rendering capabilities. The model is a single- and double-stream diffusion transformer with 17B parameters. We've also open-sourced the weights of the the accelerated version with meanflow which reduces inference steps from 100 to just 8, and PromptEnhancer, the first industrial-grade rewriting model that enhances your prompts for more nuanced and expressive image generation. Now, creators turn complex ideas—like posters with slogans or multi-panel comics—into visuals faster than ever. We’re just getting started. Stay tuned for our native multimodal image generation model coming soon. 🌐Website: 🔗Github: 🤗Hugging Face: ✨Hugging Face Demo:

Tencent Hy

89,257 просмотров • 10 месяцев назад

Super excited to introduce Pandora, a generative video World Model interactively controllable by language. #Sora and #GPT4 are both powerful. How about fusing them in a single model? 💥 Pandora gives a preview:🔭 > Build a General World Model (GWM) super efficiently by integrating pretrained autoregressive LLM and diffusion Video Model, aligning them in the representation space. > Let the LLM control the VM on-the-fly. Instruction tuning maximizes the controllability. > Autoregressive LLM empowers VM to generate indefinitely long videos: Starting with a VM for 2-second videos, Pandora extends it for 8-second videos. Would #Sora+#GPT4 under Pandora produce hours-long videos? 📽️ > World Model is beyond just video generation. It’s sensory-level information processing + concept-level reasoning and reflection. Pandora bridges both, with the concept- / language-level backbone (LLM) managing & steering the sensory-level VM functionalities. 👁️🧠 Check out for a bunch of interesting results:
1:35

Super excited to introduce Pandora, a generative video World Model interactively controllable by language. #Sora and #GPT4 are both powerful. How about fusing them in a single model? 💥 Pandora gives a preview:🔭 > Build a General World Model (GWM) super efficiently by integrating pretrained autoregressive LLM and diffusion Video Model, aligning them in the representation space. > Let the LLM control the VM on-the-fly. Instruction tuning maximizes the controllability. > Autoregressive LLM empowers VM to generate indefinitely long videos: Starting with a VM for 2-second videos, Pandora extends it for 8-second videos. Would #Sora+#GPT4 under Pandora produce hours-long videos? 📽️ > World Model is beyond just video generation. It’s sensory-level information processing + concept-level reasoning and reflection. Pandora bridges both, with the concept- / language-level backbone (LLM) managing & steering the sensory-level VM functionalities. 👁️🧠 Check out for a bunch of interesting results:

Zhiting Hu

63,456 просмотров • 2 лет назад

We have SLAM on the Robocap! 🎉 Visualized with Rerun Using NVIDIA AI Developer cuVSLAM for GPU-accelerated multicamera tracking. I basically wrote zero code myself and fully used Claude Code for this. It worked because I had so many existing examples to point to that it just wrote everything the way I would have. A few technical wins: 1. Used rattler-build from prefix.dev to package the compiled cuVSLAM CUDA binaries, which made it SUPER easy to use across repos. This also means it works on the DGX Spark (ARM64) out of the box. 2. Zero-setup experience: git clone && pixi run track-robocap auto-downloads a 100MB dataset from HuggingFace and tracks frames. 3. Real-time 3D visualization with trajectories, landmarks, pose graphs, and video playback in Rerun. Still visual-only (not visual-inertial yet), and loop closure needs some debugging. Next steps are getting this into a Gradio interface, then into daggr, and extending it to work with other datasets from exoego-forge. The last piece I'm excited about: Rerun's RRD files now support layers for incremental data. Planning to build pipelines that go from raw sensor data → slam -> human pose → depth estimation → etc. Repo here:
0:21

We have SLAM on the Robocap! 🎉 Visualized with Rerun Using NVIDIA AI Developer cuVSLAM for GPU-accelerated multicamera tracking. I basically wrote zero code myself and fully used Claude Code for this. It worked because I had so many existing examples to point to that it just wrote everything the way I would have. A few technical wins: 1. Used rattler-build from prefix.dev to package the compiled cuVSLAM CUDA binaries, which made it SUPER easy to use across repos. This also means it works on the DGX Spark (ARM64) out of the box. 2. Zero-setup experience: git clone && pixi run track-robocap auto-downloads a 100MB dataset from HuggingFace and tracks frames. 3. Real-time 3D visualization with trajectories, landmarks, pose graphs, and video playback in Rerun. Still visual-only (not visual-inertial yet), and loop closure needs some debugging. Next steps are getting this into a Gradio interface, then into daggr, and extending it to work with other datasets from exoego-forge. The last piece I'm excited about: Rerun's RRD files now support layers for incremental data. Planning to build pipelines that go from raw sensor data → slam -> human pose → depth estimation → etc. Repo here:

Pablo Vela

50,244 просмотров • 4 месяцев назад

.Beff (e/acc) explains how thermodynamic computing fits into the current computing stack: "We don't do GPUs. They're called TSUs, thermodynamic sampling units. We think people are gonna put them next to GPUs or whatever their favorite accelerator is." "You don't need to run the whole workload on the TSU, but it could run certain parts of the workload, just like the GPU is complementary to the CPU initially." "The TSU is gonna help give you more performance per watt in your data center when combined with existing build-outs." "Let's say you're serving a large diffusion model. What we showed is that by using a TSU, you can use something called an energy-based model instead of a simple probabilistic kernel." "Overall you can have basically 100 times the performance per watt of just running on pure GPUs because our chips are much lower wattage."
3:48

.Beff (e/acc) explains how thermodynamic computing fits into the current computing stack: "We don't do GPUs. They're called TSUs, thermodynamic sampling units. We think people are gonna put them next to GPUs or whatever their favorite accelerator is." "You don't need to run the whole workload on the TSU, but it could run certain parts of the workload, just like the GPU is complementary to the CPU initially." "The TSU is gonna help give you more performance per watt in your data center when combined with existing build-outs." "Let's say you're serving a large diffusion model. What we showed is that by using a TSU, you can use something called an energy-based model instead of a simple probabilistic kernel." "Overall you can have basically 100 times the performance per watt of just running on pure GPUs because our chips are much lower wattage."

MTS

42,570 просмотров • 1 месяц назад

We made a thing! Very happy to announce sqlcoder-pro and the Defog Alignment Platform. Available to use immediately without a wait-list, weights will be open-sourced very soon. The video does a quick show and tell comparison against ChatGPT (with gpt-4o). Read on for more details! TLDR 💪 equal (or better) performance on text-to-SQL as the most capable Claude-3.5 or GPT-4 models 🤝 You can use it today on a free plan/free trial, without a waitlist 🪽 self-hostable on a single RTX4090, with 2 second median generation times for SQL queries 🔁 exactly the same output every time, give the same prompt 👨🏻‍🏫 teachable and steerable: show the model what you want it to do 🛞 debuggable – you can understand WTF is going on inside the model, instead of treating it like a black box Let's dig into each of these one-by-one! Performance SQLCoder-8b-pro significantly exceeds the performance of our previous sqlcoder-8b model on Postgres text-to-SQL (from 88.2% to 90.2% accuracy - gpt-4o is at 87.6%, for reference). It is also better at following instructions. This was done via self-merges, hand crafted fine-tuning data, and adapting the training data to fit our tokenizer. Cost You can host this on the model on a single $3,500 RTX4090, and support ~5 requests/second via VLLM. If you're looking to host on the cloud instead, you can run it on a single L4 GPU that costs $300/mo on GCP Repeatability We have a dense 8b model with no MoE shenanigans. For the same prompt with temperature=0, you'll always get the same answer – which is critical in BI. Teachable In our alignment and feedback modes, you can give the model feedback on how it answered certain questions, and it will automatically adapt to the feedback. Debuggable You can use logprobs and attention scores to determine where, exactly is the model paying attention to inside a prompt + what it's getting confused by when generating outputs. Available today You can use Defog on the cloud today by going to docs[dot]defog[dot]ai, and getting an API key. Excited to hear what you think!
2:33

We made a thing! Very happy to announce sqlcoder-pro and the Defog Alignment Platform. Available to use immediately without a wait-list, weights will be open-sourced very soon. The video does a quick show and tell comparison against ChatGPT (with gpt-4o). Read on for more details! TLDR 💪 equal (or better) performance on text-to-SQL as the most capable Claude-3.5 or GPT-4 models 🤝 You can use it today on a free plan/free trial, without a waitlist 🪽 self-hostable on a single RTX4090, with 2 second median generation times for SQL queries 🔁 exactly the same output every time, give the same prompt 👨🏻‍🏫 teachable and steerable: show the model what you want it to do 🛞 debuggable – you can understand WTF is going on inside the model, instead of treating it like a black box Let's dig into each of these one-by-one! Performance SQLCoder-8b-pro significantly exceeds the performance of our previous sqlcoder-8b model on Postgres text-to-SQL (from 88.2% to 90.2% accuracy - gpt-4o is at 87.6%, for reference). It is also better at following instructions. This was done via self-merges, hand crafted fine-tuning data, and adapting the training data to fit our tokenizer. Cost You can host this on the model on a single $3,500 RTX4090, and support ~5 requests/second via VLLM. If you're looking to host on the cloud instead, you can run it on a single L4 GPU that costs $300/mo on GCP Repeatability We have a dense 8b model with no MoE shenanigans. For the same prompt with temperature=0, you'll always get the same answer – which is critical in BI. Teachable In our alignment and feedback modes, you can give the model feedback on how it answered certain questions, and it will automatically adapt to the feedback. Debuggable You can use logprobs and attention scores to determine where, exactly is the model paying attention to inside a prompt + what it's getting confused by when generating outputs. Available today You can use Defog on the cloud today by going to docs[dot]defog[dot]ai, and getting an API key. Excited to hear what you think!

Rishabh Srivastava

13,433 просмотров • 1 год назад

Researchers found a way to make LLMs 8.5x faster! (without compromising accuracy) Speculative decoding is quite an effective way to address the single-token bottleneck in traditional LLM inference. A small "draft" model first generates the next several tokens, then the large model verifies all of them at once in a single forward pass. If a token at any position is wrong, you keep everything before it and restart from there. This never does worse than normal decoding. But current drafters in Speculative decoding still guess one token at a time. That makes the drafting step itself a bottleneck, capping real-world speedups at 2-3x. DFlash is a new technique that swaps the autoregressive drafter with a lightweight block diffusion model that guesses all tokens in one parallel shot. Drafting cost stays flat no matter how many tokens you speculate. On top of that, the drafter is conditioned on hidden features pulled from multiple layers of the target model and injected into every draft layer, so it makes significantly better guesses than a drafter working from scratch. In the side-by-side demo below, vanilla decoding runs at 48.5 tokens/sec. DFlash hits 415 tokens/sec on the same model, with zero quality loss. It's already integrated with vLLM, SGLang, and Transformers, with draft models on HuggingFace for several models like Qwen3, Qwen3.5, Llama 3.1, Kimi-K2.5, gpt-oss, and many more. I have shared the GitHub repo in the replies! KV caching is another must-know technique to boost LLM inference. I recently wrote an article about it. Read it below. 👉 Over to you: What use case are you working on that can benefit from this new technique?
0:26

Researchers found a way to make LLMs 8.5x faster! (without compromising accuracy) Speculative decoding is quite an effective way to address the single-token bottleneck in traditional LLM inference. A small "draft" model first generates the next several tokens, then the large model verifies all of them at once in a single forward pass. If a token at any position is wrong, you keep everything before it and restart from there. This never does worse than normal decoding. But current drafters in Speculative decoding still guess one token at a time. That makes the drafting step itself a bottleneck, capping real-world speedups at 2-3x. DFlash is a new technique that swaps the autoregressive drafter with a lightweight block diffusion model that guesses all tokens in one parallel shot. Drafting cost stays flat no matter how many tokens you speculate. On top of that, the drafter is conditioned on hidden features pulled from multiple layers of the target model and injected into every draft layer, so it makes significantly better guesses than a drafter working from scratch. In the side-by-side demo below, vanilla decoding runs at 48.5 tokens/sec. DFlash hits 415 tokens/sec on the same model, with zero quality loss. It's already integrated with vLLM, SGLang, and Transformers, with draft models on HuggingFace for several models like Qwen3, Qwen3.5, Llama 3.1, Kimi-K2.5, gpt-oss, and many more. I have shared the GitHub repo in the replies! KV caching is another must-know technique to boost LLM inference. I recently wrote an article about it. Read it below. 👉 Over to you: What use case are you working on that can benefit from this new technique?

Avi Chawla

157,390 просмотров • 2 месяцев назад

Here is how simple it is to build a UI front-end for an AI agent (assuming you are using . The video here is 90 seconds and will show you everything you need to connect an agent to a web app: 1. Set up CopilotKit🪁 in your app 1. Copy the agent's localhost URL from LangGraph Studio 2. Paste that URL into a CopilotKit🪁 remote endpoint That's all you need to get a front-end to interact with the agent! The secret sauce here is CoAgents, an open-source library. You can use CoAgents to integrate any agent with your web application. Here is what you get: • Human-in-the-loop to steer and correct the agent • Stream intermediate agent state • Real-time state sharing between the agent and the application • Agentic generative UI to build trust that the agent is on the right path Watch the attached demo so you see it in action.
1:32

Here is how simple it is to build a UI front-end for an AI agent (assuming you are using . The video here is 90 seconds and will show you everything you need to connect an agent to a web app: 1. Set up CopilotKit🪁 in your app 1. Copy the agent's localhost URL from LangGraph Studio 2. Paste that URL into a CopilotKit🪁 remote endpoint That's all you need to get a front-end to interact with the agent! The secret sauce here is CoAgents, an open-source library. You can use CoAgents to integrate any agent with your web application. Here is what you get: • Human-in-the-loop to steer and correct the agent • Stream intermediate agent state • Real-time state sharing between the agent and the application • Agentic generative UI to build trust that the agent is on the right path Watch the attached demo so you see it in action.

Santiago

46,109 просмотров • 1 год назад

LEDGER HARDWARE WALLET COMPARISON 🚨 In this video I will show you all the current Ledger Hardware Wallets in action and tell you in the text below, which one is my favorite out of them, so you can better decide, which one to get for yourself: 1. Nano X (0:08) 2. Nano S Plus (0:52) 3. STAX (1:03) 4. FLEX (2:00) 5. STAX & FLEX side by side (2:50) You can find the different tech specs on the ledger website or feel free to ask in the comments below 👇 After using all of them, my winner is the new FLEX. The touchscreen and the NFC capabilities are a huge improvement over the normal X & S. While the STAX is the more impressive piece of tech with its curved eINK display, the lower price point of the FLEX combined with the more robust build quality, the faster responsiveness of the display and the louder speaker makes it the winner for me 🏆 Which one would you choose? 👇
3:07

LEDGER HARDWARE WALLET COMPARISON 🚨 In this video I will show you all the current Ledger Hardware Wallets in action and tell you in the text below, which one is my favorite out of them, so you can better decide, which one to get for yourself: 1. Nano X (0:08) 2. Nano S Plus (0:52) 3. STAX (1:03) 4. FLEX (2:00) 5. STAX & FLEX side by side (2:50) You can find the different tech specs on the ledger website or feel free to ask in the comments below 👇 After using all of them, my winner is the new FLEX. The touchscreen and the NFC capabilities are a huge improvement over the normal X & S. While the STAX is the more impressive piece of tech with its curved eINK display, the lower price point of the FLEX combined with the more robust build quality, the faster responsiveness of the display and the louder speaker makes it the winner for me 🏆 Which one would you choose? 👇

checker4e

22,896 просмотров • 1 год назад

This is probably the most complex workflow I’ve ever built, only with open-source tools. It took my 4 days. It takes four inputs: author, title, and style; and generates a full visual animated story in one click in ComfyUI . I worked on it for four days. There are still some bugs, but here’s the first preview. Here’s a quick breakdown: - The four inputs are sent to LLMs with precise instructions to generate: first, prompts for images and image modifications; second, prompts for animations; third, prompts for generating music. - All voices are generated from the text and timed precisely, as they determine the length of each animation segment. - The first image and video are generated to serve as the title, but also as the guide for all other images created for the video. - Titles and subtitles are also added automatically in Comfy. - I also developed a lot of custom nodes for minor frame calculations, mostly to match audio and video. - The full system is a large loop that, for each line of text, generates an image and then a video from that image. The loop was the hardest part to build in this workflow, so it can process either a 20-second video or a 2-minute video with the same input. - There are multiple combinations of LLMs that try to understand the text in the best way to provide the best prompts for images and video. - The final video is assembled entirely within ComfyUI. - The music is generated based on the LLM output and matches the exact timing of the full animation. - Done! For reference, this workflow uses a lot of models and only works on an RTX 6000 Pro with plenty of RAM. My goal is not to replace humans, as I’ll try to explain later, this workflow is highly controlled and can be adapted or reworked at any point by real artists! My aim was to create a tool that can animate text in one go, allowing the AI some freedom while keeping a strict flow. I don’t know yet how I’ll share this workflow with people, I still need to polish it properly, but maybe through Patreon. Anyway, I hope you enjoy my research, and let’s always keep pushing further! :)
0:55

This is probably the most complex workflow I’ve ever built, only with open-source tools. It took my 4 days. It takes four inputs: author, title, and style; and generates a full visual animated story in one click in ComfyUI . I worked on it for four days. There are still some bugs, but here’s the first preview. Here’s a quick breakdown: - The four inputs are sent to LLMs with precise instructions to generate: first, prompts for images and image modifications; second, prompts for animations; third, prompts for generating music. - All voices are generated from the text and timed precisely, as they determine the length of each animation segment. - The first image and video are generated to serve as the title, but also as the guide for all other images created for the video. - Titles and subtitles are also added automatically in Comfy. - I also developed a lot of custom nodes for minor frame calculations, mostly to match audio and video. - The full system is a large loop that, for each line of text, generates an image and then a video from that image. The loop was the hardest part to build in this workflow, so it can process either a 20-second video or a 2-minute video with the same input. - There are multiple combinations of LLMs that try to understand the text in the best way to provide the best prompts for images and video. - The final video is assembled entirely within ComfyUI. - The music is generated based on the LLM output and matches the exact timing of the full animation. - Done! For reference, this workflow uses a lot of models and only works on an RTX 6000 Pro with plenty of RAM. My goal is not to replace humans, as I’ll try to explain later, this workflow is highly controlled and can be adapted or reworked at any point by real artists! My aim was to create a tool that can animate text in one go, allowing the AI some freedom while keeping a strict flow. I don’t know yet how I’ll share this workflow with people, I still need to polish it properly, but maybe through Patreon. Anyway, I hope you enjoy my research, and let’s always keep pushing further! :)

Lovis Odin

58,571 просмотров • 9 месяцев назад

HOW TO SELL "SECOND BRAIN AS A SERVICE" FOR $5K + (FULL COURSE) The model: build businesses a structured knowledge base out of plain markdown files, then charge to maintain it. No fancy database, Obsidian, or RAG. It becomes the foundation for every other AI project you sell them afterward. A $750 audit turns into a $3,500 build turns into a client who keeps paying you to add more. Adam Sandler came on the pod and walked through the entire playbook live. Here's what I learned: 1. Lead with the knowledge base, not the agent. Every prospect wants a CFO agent. Nobody has the context to make one work. Solve "where does the info live?" first and you've teed up every future build. 2. Sell it with AI out of the pitch. "I'll organize all your scattered company knowledge into one living asset." Their files are everywhere and they hate it. That closes on its own. 3. Zero technical friction. Just markdown files on their machine. No signup, no vault, no vectors. Technical drop-off kills AI deals. This deletes it. 4. Seven note types are your starting schema. Snapshot, people, rules, project history, decisions, open loops, links. Walk in with this and you already have a point of view. 5. Find the spine. The one thing everything ladders up to. Usually their annual goals. 6. Portability is the pitch. Fable got pulled and everyone panicked. A knowledge base outside Anthropic or OpenAI plugs into any model. Sells itself. 7. Token savings close the enterprise. One source means you stop re-pasting context every session. Real money, especially with pay-as-you-go API pricing. 8. The audit is the tripwire. Charge to map where their knowledge lives, then roll the fee into the build. Half the build is already done. 9. Two skills run it on autopilot. Ingest pulls from Gmail, calendar, and CRM. Curate does a weekly cleanup pass. That's the flywheel. 10. The knowledge base tells you what to sell next. Ask it for the top 3 AI opportunities. It finds the gap ("inbound takes 22 hours") and the next build sells itself. His 2 key takeaways: 1. Build one for yourself first. One source of truth, built once, every tool connected. Live the unlock and you'll sell it with conviction. 2. Don't let startups scare you off. Off-the-shelf can't learn how your client's business works. That domain knowledge is the moat you charge for. Adam is quietly building one of the best AI service offers out there and we had a blast going deep on it. Go follow Adam Sandler Full video below. (Also available on the Build With AI podcast wherever you get your pods)
53:21

HOW TO SELL "SECOND BRAIN AS A SERVICE" FOR $5K + (FULL COURSE) The model: build businesses a structured knowledge base out of plain markdown files, then charge to maintain it. No fancy database, Obsidian, or RAG. It becomes the foundation for every other AI project you sell them afterward. A $750 audit turns into a $3,500 build turns into a client who keeps paying you to add more. Adam Sandler came on the pod and walked through the entire playbook live. Here's what I learned: 1. Lead with the knowledge base, not the agent. Every prospect wants a CFO agent. Nobody has the context to make one work. Solve "where does the info live?" first and you've teed up every future build. 2. Sell it with AI out of the pitch. "I'll organize all your scattered company knowledge into one living asset." Their files are everywhere and they hate it. That closes on its own. 3. Zero technical friction. Just markdown files on their machine. No signup, no vault, no vectors. Technical drop-off kills AI deals. This deletes it. 4. Seven note types are your starting schema. Snapshot, people, rules, project history, decisions, open loops, links. Walk in with this and you already have a point of view. 5. Find the spine. The one thing everything ladders up to. Usually their annual goals. 6. Portability is the pitch. Fable got pulled and everyone panicked. A knowledge base outside Anthropic or OpenAI plugs into any model. Sells itself. 7. Token savings close the enterprise. One source means you stop re-pasting context every session. Real money, especially with pay-as-you-go API pricing. 8. The audit is the tripwire. Charge to map where their knowledge lives, then roll the fee into the build. Half the build is already done. 9. Two skills run it on autopilot. Ingest pulls from Gmail, calendar, and CRM. Curate does a weekly cleanup pass. That's the flywheel. 10. The knowledge base tells you what to sell next. Ask it for the top 3 AI opportunities. It finds the gap ("inbound takes 22 hours") and the next build sells itself. His 2 key takeaways: 1. Build one for yourself first. One source of truth, built once, every tool connected. Live the unlock and you'll sell it with conviction. 2. Don't let startups scare you off. Off-the-shelf can't learn how your client's business works. That domain knowledge is the moat you charge for. Adam is quietly building one of the best AI service offers out there and we had a blast going deep on it. Go follow Adam Sandler Full video below. (Also available on the Build With AI podcast wherever you get your pods)

Corey Ganim

105,458 просмотров • 19 дней назад

Tune Studio is an end-to-end platform for developing applications using Large Language Models. So far, I haven't seen any other platform like this one. You can do everything here: 1. You can curate your data. 2. Use the playground to play with different models and try your ideas. 3. Fine-tune an open-source model on your data. 4. Deploy the model when you are done. This is awesome for anyone building generative AI applications. You can use Tune Studio to work with any of the open-source models out there. They were one of the few companies to host Llama 2 and Llama 3 before anyone else. Here is a link to check it out: One of their main selling points is that Tune Studio scales! You don't have to worry about serving your model to lots of users. They also have built-in user management, authentication, on-prem support, user context management, and pretty much everything you need to build generative AI applications. Thanks to the Tune team for collaborating with me on this post. We are living through the best years of development tools for AI developers. The field is unstoppable.
0:23

Tune Studio is an end-to-end platform for developing applications using Large Language Models. So far, I haven't seen any other platform like this one. You can do everything here: 1. You can curate your data. 2. Use the playground to play with different models and try your ideas. 3. Fine-tune an open-source model on your data. 4. Deploy the model when you are done. This is awesome for anyone building generative AI applications. You can use Tune Studio to work with any of the open-source models out there. They were one of the few companies to host Llama 2 and Llama 3 before anyone else. Here is a link to check it out: One of their main selling points is that Tune Studio scales! You don't have to worry about serving your model to lots of users. They also have built-in user management, authentication, on-prem support, user context management, and pretty much everything you need to build generative AI applications. Thanks to the Tune team for collaborating with me on this post. We are living through the best years of development tools for AI developers. The field is unstoppable.

Santiago

39,101 просмотров • 2 лет назад