正在加载视频...

视频加载失败

Want better MiniMax-H3 LoRAs? DiffSynth-Studio just open-sourced two things worth checking out. 🔥 ⚙️ The MiniMax-H3 Training Adapter, a rank 64 LoRA. Combined with differential training, it helps you train higher quality LoRAs: 🔗 📦 The self-generated dataset behind the adapter, fully open as well: 🔗 One more thing:...

10,045 次观看 • 5 天前 •via X (Twitter)

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

We are in an insane run of open-weight drops. Every modality, open source is winning. This is what an open source AI summer ☀️ looks like: 🧠 LLMs & Reasoning → DeepSeek-V4-Flash-0731 (my king 👑): 304B MoE refresh, Terminal-Bench 2.1 jumps 61.8→82.7 over the preview, DeepSWE 7.3→54.4. Closes in on Opus-4.8 on Agents' Last Exam (25.2 vs 25.7). MIT. → Muse-Glimmer-30B, from Meta (they are back!!): their first open agentic model. ~29.6B dense + perception encoder, 131k+ context, built to run fully local, no cloud. Apache 2.0. → Liquid AI LFM2.5-2.6B: 2.69B params, 131k context, 220 tok/s on an M5 Max in under 2.5GB RAM. Competitive with models 4x larger on agentic tasks. → inclusionAI Ling-3.0-flash: 124B total, only 5.1B active, ~12% the size of their old 1T flagship Ring-2.6, matches it on key benchmarks. MIT. → inclusionAI Ling-3.0-tiny: 7.9B total, 1.3B active, 86-90 tok/s on an M4 Pro MacBook at ~8GB peak memory. MIT. → NVIDIA Nemotron-3.5-Lightning-30B-A3B: hybrid Mamba-2+MoE+Attention, up to 1M context, runs on a single H100 or DGX Spark, SWE-bench Verified 52.8. → deepgrove maple-preview: 20B-A1B ternary-weight reasoner, 218 tok/s on a Mac mini M4, 5.3GB checkpoint. MIT. → BigBang-v1 (endless-frontier): fine-tuned from Qwen3.6-35B-A3B via a self-evolving generator/critic synthetic-data loop. Lands aggregate performance between DeepSeek V4 Flash (284B) and V4 Pro (1.6T), at 35B. Apache 2.0. 🎬 Video → MiniMax-H3: 33B dense omni model, native stereo audio, up to 2K/15s. 3.6k+ likes already. → Minimax-H3-Turbo (lightx2v): Apache-2.0 turbo distillation of H3 for fast inference. → Lightricks LTX-2.5: image-to-video update, custom Gemma-4-12B text encoder, a markedly stronger distilled model. 🔊 Voice → NVIDIA NemotronLabs VoiceChat-11B: full-duplex speech-to-speech, ~450ms turn-taking, #2 on open VoiceBench, and the first open full-duplex model with live tool-calling mid-conversation. 🛡️ Safety → Mistral Shieldstral-1.0-3B: 3B multimodal guardrail that takes your safety policy as plain text instead of fixed categories. Beats LlamaGuard-4-12B and ShieldGemma-9B on HarmBench (99.4) and ToxicChat (84.1) at a fraction of the size. Apache 2.0.

Victor M

54,264 次观看 • 17 天前

Introducing ml-intern, the agent that just automated the post-training team Hugging Face It's an open-source implementation of the real research loop that our ML researchers do every day. You give it a prompt, it researches papers, goes through citations, implements ideas in GPU sandboxes, iterates and builds deeply research-backed models for any use case. All built on the Hugging Face ecosystem. It can pull off crazy things: We made it train the best model for scientific reasoning. It went through citations from the official benchmark paper. Found OpenScience and NemoTron-CrossThink, added 7 difficulty-filtered dataset variants from ARC/SciQ/MMLU, and ran 12 SFT runs on Qwen3-1.7B. This pushed the score 10% → 32% on GPQA in under 10h. Claude Code's best: 22.99%. In healthcare settings it inspected available datasets, concluded they were too low quality, and wrote a script to generate 1100 synthetic data points from scratch for emergencies, hedging, multilingual etc. Then upsampled 50x for training. Beat Codex on HealthBench by 60%. For competitive mathematics, it wrote a full GRPO script, launched training with A100 GPUs on watched rewards claim and then collapse, and ran ablations until it succeeded. All fully backed by papers, autonomously. How it works? ml-intern makes full use of the HF ecosystem: - finds papers on arxiv and reads them fully, walks citation graphs, pulls datasets referenced in methodology sections and on - browses the Hub, reads recent docs, inspects datasets and reformats them before training so it doesn't waste GPU hours on bad data - launches training jobs on HF Jobs if no local GPUs are available, monitors runs, reads its own eval outputs, diagnoses failures, retrains ml-intern deeply embodies how researchers work and think. It knows how data should look like and what good models feel like. Releasing it today as a CLI and a web app you can use from your phone/desktop. CLI: Web + mobile: And the best part? We also provisioned 1k$ GPU resources and Anthropic credits for the quickest among you to use.

Aksel

1,267,178 次观看 • 4 个月前

The future of housework just leaked on GitHub and nobody is talking about it. knox byte just open sourced a framework that coordinates swarms of Unitree G1 humanoid robots to clean your entire house on their own. It's called ARGOS. You tell it "clean the bedroom" in plain English and 2+ G1 robots split the room into zones, sweep in parallel, and sync up for the tasks that need four hands like making the bed or moving furniture. The Claude API decomposes your sentence into a task graph. An auction system makes every robot bid on every task based on distance, battery, and current load. The cheapest robot wins. Cooperative jobs go to the cheapest team. Here's what makes this different from every demo video Boston Dynamics keeps teasing: → 12 cleaning tasks baked in sweeping, mopping, wiping, vacuuming, taking out trash, making the bed, changing sheets, moving furniture, sorting items → 3 policy architectures running underneath OpenVLA-7B for language tasks, Diffusion Policy for floor coverage, ACT for dexterous bimanual work → Train it on your own footage record yourself cleaning, run one command, it extracts poses, builds a LeRobot dataset, and LoRA fine-tunes the policy → PEFA protocol for cooperative work Propose, Execute, Feedback, Adjust. If one robot fails halfway through making the bed, the team replans and retries → Full MuJoCo simulation so you test policies before pushing them to real hardware → Silver and cyan terminal dashboard that shows live fleet status, zone maps, task queues, and battery levels in real time The G1 robots talk to each other over CycloneDDS mesh using Unitree's native SDK. No cloud. No middleware. The whole thing runs on a Jetson Orin inside each robot. The wildest part is the training pipeline. Drop cleaning videos into a folder, run argos train ingest, and the framework does the entire pipeline frame extraction, pose estimation, action labeling, HDF5 dataset, fine-tune, evaluate in sim, deploy to robot. One command per stage. Unitree G1s already exist. The framework to make them clean your house just hit GitHub. 52 stars. MIT License. 100% Opensource.

Guri Singh

27,404 次观看 • 3 个月前

more frontend vibecoding tips (results below): WHY YOUR VIBECODED FRONTENDS ALL LOOK THE SAME AND SUCK: when asked to make a frontend, the agent/llm will default to the center/average of its training data (in a very loose sense). through the training process, the model essentially converges on some default UI style. it's very capable of doing things that are different from this style, but you have to ask! for instance, ChatGPT tends to reply in the same tone for all users untill you interact with it and instruct it differently ("be sassy", "eli5"). the second reason is that most of us are not good at coming up with designs and describing them precisely (see my tweet on a crash course in common components, which i'll link below). treat frontend generation just like any other eng task! you need to provide a good detailed spec. TIPS: 1. give ur agent screenshots of designs you like (you may not know the right words to describe them but the agent will! a pic = 1000 words) where to find ui inspo? Behance, Dribbble, Mobbin (Mobbin is paid but worth it!) 2. ask ur agent for proposals, this helps "seed" different directions so the final frontend stands out. don't be afraid to go back and forth. 3. ban certain tendencies: no Inter/Roboto, no shadcn (controversial), no gradients, no emojis 4. encourage the agent to be extreme and make bold decisions, not safe ones. i think that the underlying models tend to get taught during RL/fine-tuning to make conservative choices that produce reasonable but boring frontends 5. give ur agent Figma MCP. the best results will come if you mockup your vision in Figma first. 6. Ideally choose an agent with vision capabilities TLDR: Most people are tremendously underusing agents for frontend design. They are much better than you might expect.

andrew gao

64,712 次观看 • 6 个月前

AI token usage is up 10x in 7 months, compounding 40%/MONTH! There is NO BUBBLE when demand is STILL accelerating And this is just OpenRouter, it doesn't count the labs direct token usage and APIs But here's what's interesting about these numbers, the demand is coming from everywhere at once US models (OpenAI, Anthropic, Google) keep growing, while Chinese open weight models (DeepSeek, Tencent, Xiaomi, Minimax) grew even faster and now drive over 60% of usage on OpenRouter Closed source and open source both compounding at the same time. This is literally the best case scenario for AI Infra investors It means both frontier model tokens and cheaper tokens have product market fit. This means the application layer is finding ways to use both and generate ROI with both types Demand for tokens IS demand for compute. This is why SpaceX is looking to build 10GW of compute by next year, because the demand is clearly here Now combine this demand set up, with NVIDIA yesterday announcing financing platforms with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to mobilize over $500 billion of third party capital for AI infrastructure And Jensen has said publicly he expects $3 to $4 TRILLION of AI infrastructure spend by 2030 The build out will have to continue for a lot longer than the market is expecting, that is very clear to me. Don't let this consolidation period in AI infra stocks shake you out, they will have their moment again and take their next leg higher p.s. if you want to see how im investing in this, you can track my real-time portfolio and the research of all 5 Milk Road PRO analysts with live trade notifications, and it's just $1 to try it out (insane price just to check it out). Learn more here: Good luck out there!

Kyle Reidhead | Milk Road

28,320 次观看 • 18 天前

🚀 Introducing EgoExo Forge - built on top of Rerun, Gradio, and Hugging Face hub (I’ll be in San Francisco July 21–29 — if you’re into robotics, egocentric AI, large-scale data collection, or just want to chat, DM me!) In my opinion, large-scale, diverse, and high-quality data is still the largest bottleneck for generalized robotics deployment. I believe that some version of imitation learning from human examples will be the most scalable + clean way to train humanoid robots 🤖 (similar to what Tesla did for Full Self Driving). Teleop is too expensive to collect a large enough dataset in a reasonable manner, so passive collection via egocentric (and in certain cases, exocentric) views feels like the right bet. Over the past few months, I've been trying to build out the scaffolding for this and using Rerun as my underlying infrastructure. Data being collected needs to be easily inspectable + time series and rerun provides the right tooling for this. My goal is to first build out a ground truth representative dataset from already existing open source data, generate some reasonable baselines, and then go out and collect my own data that adheres to the defined schema. 🔍 Starting with open-source datasets 1. EgoDex from Apple 2. HOCap from Nvidia and the University of Texas at Dallas 3. Assembly101 from Meta All these different datasets have different sensor configurations + annotations, so my goal with egoexo-forge is to have one consistent labeling scheme + data layout. I built a data pipeline that aligns all of the different datasets in one general schema assuming the COCO133 keypoint layout that allows for exo+ego, ego only, or exo only Since the scaffolding is already there, it becomes MUCH easier to add other datasets. So the next ones that I'll be including are HD-EPIC kitchens dataset, HOT3D, and finally my own personal iPhone + insta360 go collection method. Once I have a diverse variety of datasets, I'll double down on what I believe to be the key algorithms required to make useful data for imitation learning 📊 1. Camera Pose estimation via SLAM/SFM for ego perspective (and automatic calibration for exo) 2. Human pose estimation for both egocentric + exocentric views 3. Metric 3D reconstruction + object tracking I'll be setting up reasonable open-source baselines for each of these to validate that these datasets work, and then finally try to use the generated datasets for some imitation learning via the pi0-lerobot repo I've been working on. I plan on making a blog post + providing more info on all of this in the near future so stay tuned

Pablo Vela

32,085 次观看 • 1 年前

A Letter to Our Community: The Road Ahead for Robotics To our Community and Partners, As we step into 2026, our mission at Axis is clearer than ever: Constructing the definitive End-to-End Scaling Layer for Robotics. Our goal is to accelerate the transfer of diverse human intelligence into Robotics General Intelligence (RGI). By owning the critical path of intelligence creation, we are turning the physical limitations of robotics into a scalable, software-driven future. Here is our strategic outlook and roadmap for the year ahead. The Core Thesis: Simulation is the Only Way Out The path to RGI is currently blocked by Data Scarcity, Generalization Fragility, and Hardware Fragmentation. At Axis, we believe Simulation is the only way out. Our Simulation Data Platform and Data Augmentation Engine transform raw data into "Synthetic Gold". Backed by academic milestones like Roboverse, Skill Blending, and GraspVLA, we have proven that pure simulation can achieve the generalization required for the real world. We don’t just collect data; we architect it. The Engine: Why Crypto? We believe RGI should come from all, not a few. Crypto is not just a feature; it is the primitive that powers our entire ecosystem flywheel: - Incentive Mechanism: Democratizing contribution and rewarding the trainers and developers. - Assetization: Turning proprietary data and refined models into liquid, ownable assets. - Verifiable Workflow: We are opening the "Black Box" of AI. By bringing total transparency to the Task Generation → Data Collection → Model Training pipeline, we ensure every byte of intelligence is verifiable, traceable, and secure. 2026 Strategic Deliverables This year, we are committed to delivering three foundational pillars: - The World's Largest Training Dataset for Robots: A robot training set—diverse, high-quality interaction data at an unprecedented scale. - A Robotics Foundation Model: A universal robotic brain trained on our pure simulation and synthetic data, capable of robust cross-embodiment transfer and open-world adaptability. - Evolvable Robot Hardware: Robots deployed with Axis models that autonomously evolve through continuous interaction, turning every deployment into a self-improving node within our RGI network. The Ultimate Vision We are building more than models; we are architecting the Distributed Machine Economy. A future where every dataset, model, and robotic embodiment is a verifiable asset in a global, autonomous network. Thank you for building the future of intelligence with us✌️📷

Axis Robotics

27,858 次观看 • 7 个月前

Are you safer with LIDAR, or are you safer with vision? This is a false dichotomy. The more pertinent question today is "do you have something, or do you have nothing?" As you can see from the clips below, vision based systems avoid countless potential collisions every day. The difference between a crash and no crash isn't what sensor suite you chose — it's whether you have any AI on your car at all. Even if we concede that LIDAR may help prevent some additional crashes, we are really debating whether it is 1% of crashes or 0.00001% of crashes. Not all crashes are super complex and require lasers to detect. Most are simple, routine, and can easily be prevented by today's vision based AI. In fact, evidence is mounting that computer vision based systems can actually outperform more traditional approaches to self-driving. Why? Because the low cost of cameras enables you to create a much larger, more varied, and more diverse dataset. If you want to have expensive custom cars that's fine, but you're going to get fewer vehicles for the same budget. Seeing what's in front of you now is actually less important than predicting what's going to happen next — and the large scale datasets used to train pure vision systems are the best for predicting what's next. Counter-intuitively, the simpler and lower cost sensor actually has properties that make it better suited for training advanced AI. Computer vision based self-driving is often framed by LIDAR proponents as "cheaping out" on the sensor suite to save money. But it's not about being cheap, it's about bringing the technology to everyone. 1.2 million people die on the road every year around the world. That's around 39 million people who've died on the roads around the world since I was born — equivalent to a city the size of Tokyo or New Delhi getting wiped off the map. The status quo is simply unacceptable, and something has to be done to fix it as soon as possible. Of the 1.2 million people that will die on the roads this year, about 40,000 will be Americans. That's about 3%. So if we moved entirely to self-driving cars in America and brought crashes down to 0, 97% of the world's crash fatalities would still be taking place as usual. Deploying a $200,000+ retrofitted self-driving car may work in a few American cities, but it is not going to make sense in most places around the world where fares are much cheaper. Most often, the choice is not between LIDAR and vision. It's between vision or nothing. The best system is the system that's there running on my car when I need it to save my life. To say that all self-driving cars must have LIDAR is to sentence most of the world to death. We can't write off computer vision if we want to make a serious dent in this problem. It's going to be a key piece of the solution. Let LIDAR based players build the best self-driving car they can, and let vision based players do the same. We need to be trying everything

Whole Mars Catalog

45,801 次观看 • 1 年前

If anyone tells you that you need to cut out carbs to lose weight… Run and never listen to anything that person says again ⁠⁠ I’ve gotten to single digit body fat numerous times while eating 2-4 slices of sourdough bread and 1-2 bagels every day every prep Here’s how I do it and how my clients do it too:⁠ ⁠ -We either track macros or build a meal plan around their lifestyle and schedule⁠ ⁠ -We build meals around foods we enjoy⁠ If I make you eat fish and broccoli, and you fucking hate fish and broccoli…I’m setting you up for failure right off the get go. Would never make any client eat something they despise when there are so many good, clean foods out there to eat to get us the results we’re after ⁠ -We stay in a slight calorie deficit without starving ourselves⁠ -IF our schedule allows it, we always try to move more or increase cardio before dropping food to keep food as high as possible for as long as possible More food = more fuel = better training sessions = helps preserve and or build muscle = better metabolism long term ⁠ - We train hard, lift heavy as fuck, and stay consistent because we’re not burned out or miserable with our food⁠ ⁠ Bread or carbs didn’t make you fat and DOES NOT make you fat ⁠ Being inconsistent and over eating is what did ⁠ If you keep jumping from diet to diet, cutting more and more food to “get results”…⁠ ⁠ You’ll never stay lean ⁠ Because you’ll never stick with it long enough for it to work because you’ll feel like death from not eating enough ⁠ But when you learn how to lose fat the right way with structure, flexibility, and a plan that fits your life…⁠ ⁠ You stop starting over and hating the process while getting lean eating the foods you actually enjoy eating On top of looking AND feeling better There’s a right way and there’s a wrong way If you wanna see how it’s done the right way, shoot me a message and I’ll show you how my team can help you as well

Bailey Schober | Men’s Fitness & Nutrition Coach

25,757 次观看 • 1 个月前

Claude CodeにMiniMax H3のプロンプトを丸投げしたら、この動画ができました😃 Claude Codeに投げたプロンプト↓ ------------------- MiniMax H3を使ってモーショングラフィックスをバリバリ使った15秒のCMを作りたい。 どんな内容が良いと思う? 条件は以下 ・モーショングラフィックスを多用 ・人物も登場させる(実写・アニメーションは問わない) ・アスペクト比は16:9 ------------------- すると、Claude Codeが3つCM案を提案。 そのうちの1つを選んだら↓のプロンプトができました。 MiniMax H3を利用したツールはMiniMax Design。 #MiniMaxH3 on #MiniMaxDesign Claude Codeが作ったMiniMax H3向けのプロンプト↓ ---------------------- integrated_multimodal_description: 2D-animated, a bold graphic-design motion-graphics commercial in 16:9 with exactly 13 distinct cuts at 24fps. This film is 85% kinetic graphic design and 15% character action: giant kinetic Japanese typography, hard-edged flat shapes, split screens, halftone, speed lines, torn-paper reveals and shutter flashes. Palette, strictly four colors: fluorescent orange, deep black, off-white, silver grey. The recurring character is a young anime street dancer with a black chin-length bob and blunt bangs, sharp confident eyes, an oversized off-white track jacket with fluorescent-orange sleeve stripes, black cargo shorts, and chunky white sneakers with fluorescent-orange lines. Preserve her exact face, proportions, hairstyle, outfit, materials and colors throughout; never redesign her. Every graphic element moves fast and snaps hard on the beat. [Shot 1] On a deep black field, a massive off-white circle slams into frame left of center, then two fluorescent-orange bars wipe across the upper and lower thirds, then the Japanese kanji "烈" builds itself stroke by stroke in off-white at the center until it is complete. The camera shakes slightly with small amplitude at fast speed on each impact. [Shot 2] At 00:01.000, the camera cuts to an off-white field where four black squares rotate into the center one at a time. The largest square opens as a porthole revealing an extreme close-up of the dancer's ice-sharp eye glancing up, then a red-and-cyan split flash tears across the edges and the squares shatter into flat shards. The camera pushes in with small amplitude at fast speed. [Shot 3] At 00:02.000, the camera cuts to an off-white field filled by an enormous black kanji "烈". The dancer sprints in from frame left, power-slides along the baseline of the typography, and fluorescent-orange ink splatter trails behind her as speed lines streak past. The camera trucks left with large amplitude at fast speed. The kanji stays fully readable before she crosses in front of it. [Shot 4] At 00:03.100, the camera cuts to a vertical triptych of fluorescent-orange, deep black and off-white panels. Her flat silhouette appears in each panel in stroboscopic freeze, one after another — takeoff, mid-air tuck, landing — while giant vertical silver text "全速" scrolls upward behind the panels. Static shot. [Shot 5] At 00:04.000, the camera cuts to a black field where a giant tilted ring of fluorescent-orange Japanese characters reading "止マルナ" rotates. The dancer leaps through the center of the ring, spins once in mid-air, and the characters break off and scatter outward as flat shards. The camera arcs with large amplitude at fast speed, then snap-zooms onto her confident face. [Shot 6] At 00:05.100, the camera cuts to a clean off-white editorial card holding a huge black kanji "跳" with one orange diagonal slash. She vaults over the word itself, plants her palm on its left radical, the character compresses like a spring under her hand, then rebounds as her legs whip across the frame. Static shot. [Shot 7] At 00:06.000, the camera cuts to a full fluorescent-orange field crossed by one thick black diagonal stripe. She back-flips along the stripe in three stroboscopic ghost frames tinted black, off-white and silver grey, while a large outlined kanji "回" rotates 180 degrees in sync with her rotation. The camera rolls clockwise with small amplitude at fast speed. [Shot 8] At 00:07.000, the camera cuts to an off-white field with a hard black band. Four bold black kanji slam in one at a time at different scales and angles — "速", then "鋭", then "静", then "烈" — each landing with a white shutter flash and an orange impact mark, while she slides on her knees across the foreground, jacket flaring. The camera shakes slightly with small amplitude at fast speed. [Shot 9] At 00:08.000, the camera cuts to a black frame where a white wireframe grid tilts away in 3D. She runs up the grid like a wall, kicks off it, and hangs frozen in mid-air as a fluorescent-orange circle stamps around her pose like a target lock with silver UI brackets and tick marks. The camera tilts up with small amplitude at slow speed. [Shot 10] At 00:09.000, the freeze releases and the camera cuts to layered flat-color panes stacked toward the lens. She dives straight at the camera, the panes tear open one at a time like thick paper, and each tear reveals a larger portion of the black kanji "烈" behind, before her sneaker wipes across the foreground. The camera pulls out with large amplitude at fast speed. [Shot 11] At 00:10.100, the camera cuts to a rapid-fire montage of four full-screen graphic posters of her in different action poses — mid-flip, sliding, landing, hands in pockets — snapping past with hard cuts, each with an oversized silver number 01 to 04, a black barcode strip and thick diagonal slashes. Static shot. [Shot 12] At 00:11.100, the camera cuts to a clean off-white cyclorama for the hero moment. She lands a final backflip dead center in slow motion, straightens with both hands in her pockets, and a shockwave of concentric fluorescent-orange rings blasts outward from her feet carrying shattered fragments of the kanji "烈" with it, then the frame overexposes toward white. The camera pushes in with small amplitude at slow speed. [Shot 13] At 00:13.000, the camera cuts to the final identity card on a pale off-white field. An enormous black kanji "烈" dominates the frame with translucent orange rings and thin technical arcs behind it, the silver-grey word "RETSU" sits directly beneath the kanji, and she stands relaxed at the lower center overlapping the bottom of the strokes with her jacket rippling. One thin orange pulse sweeps horizontally through the typography and holds cleanly until exactly 15.000 seconds. Static shot. Editing: extremely aggressive rhythm — hard cuts on every beat, graphic matches, whip pans, snap zooms, stroboscopic freezes, foreground wipes, split flashes and impact shakes. Every cut must feel compositionally different. All Japanese typography is always fully readable before the character overlaps it. No weapons, no combat, no fire, no explosions; all energy comes from motion design, ink, torn paper, wind and her street-dance athleticism. No photorealism, no 3D rendering, no gradients, no browser interface, no player controls, no subtitles, no captions, and no text other than the Japanese characters named above. Exactly 13 cuts. overall_soundscape: Sneaker impacts, knee slides and landings hit hard against a clean room tone. Ink splatters burst, thick paper tears open, and flat panels shatter like glass. Sharp whooshes carry the color wipes and typography slams, with small shutter clicks on each flash. No dialogue. non_diegetic_music: An original 15-second drum-heavy breakbeat with a deep sub bass, hard snare hits, risers and glitch fills locked to every cut. Begin with a single impact, build tightly through the middle, peak at the landing at 00:11.100, and end with a cold electronic logo sting that resolves at 15.000 seconds.

タナベ | AI動画 × マーケティング

55,956 次观看 • 9 天前