Slava 🍁❤️ 🇺🇦 @pmbstuff

Web Engineer Canada Joined July 2010

Tweets

471
Followers

146
Following

103
Likes

19

SkalskiP @skalskip92

4 days ago

supervision, the open-source library I created almost 3 years ago, is crossing 35,000 stars on GitHub! supervision is the engine behind all my demos, including basketball AI. link: github.com/roboflow/super…

28 189 2K 117K 2K

Download Video

Haocheng Xi @HaochengXiUCB

4 days ago

🚀 Introducing Sparse VideoGen2 (SVG2) — Pareto-frontier video generation acceleration with semantic-aware sparse attention! 🏆Spotlight paper accepted by #NeurIPS2025 ✅ Training-free & plug-and-play ✅ Up to 2.5× faster on HunyuanVideo, 1.9× faster on Wan 2.1 ✅ SOTA quality…

15 54 249 34K 147

Download Video

Google AI Developers @googleaidevs

5 days ago

Google AI Pro and Ultra subscribers now have access to Gemini Code Assist and Gemini CLI with higher daily limits. Get AI agent and coding assistance powered by Gemini 2.5 right in your IDE and terminal: blog.google/technology/dev…

75 232 1K 224K 346

Chrome for Developers @ChromiumDev

6 days ago

Announcing our public preview of Chrome DevTools MCP! Experience the full power of DevTools in your AI coding agent → goo.gle/4pDE6Tk With Chrome DevTools MCP, your AI agent can run performance traces, inspect the DOM, & perform real-time debugging of your web pages.

109 726 4K 988K 3K

Download Video

Tongyi Lab @Ali_TongyiLab

6 days ago

Announcing the open-source release of Qwen3-VL! A powerful vision-language model that can operate GUIs, code Draw.io charts from mockups, and recognize "everything" from daily life to specialized fields. Highlights: 🔹 Precise event location in videos up to 2…

23 179 1K 58K 483

Download Video

Gabriel Synnaeve @syhw

5 days ago

(🧵) Today, we release Meta Code World Model (CWM), a 32-billion-parameter dense LLM that enables novel research on improving code generation through agentic reasoning and planning with world models. ai.meta.com/research/publi…

59 299 2K 788K 1K

Chubby♨️ @kimmonismus

7 days ago

Lets go!

41 36 1K 79K 90

Download Image

Chubby♨️ @kimmonismus

2 weeks ago

I did not see that coming: Chain of Thought in AI Video generation! LumaAI introduced Ray 3 - HDR Generation - Fidelity "state of the art realism, physics, character consistency, and far superior instruction following" - SOTA Physics & Controls And much more. Some examples:

Luma AI @LumaLabsAI

2 weeks ago

12 23 263 115K 136

Download Video

38 82 853 68K 325

Download Video

elhacker.NET @elhackernet

2 weeks ago

448 1K 7K 956K 2K

Download Video

Min Choi @minchoi

2 weeks ago

This ChatGPT prompt literally makes ChatGPT give clear, accurate, and to the point answers

52 75 776 139K 1K

Download Image

OpenAI @OpenAI

2 weeks ago

We’re releasing GPT-5-Codex — a version of GPT-5 further optimized for agentic coding in Codex. Available in the Codex CLI, IDE Extension, web, mobile, and for code reviews in Github. openai.com/index/introduc…

471 1K 8K 2.9M 2K

Tom Dörr @tom_doerr

2 weeks ago

generate unit tests for your code using LLMs

6 114 886 44K 967

Download Image

🔥 Matt Dancho (Business Science) 🔥 @mdancho84

2 weeks ago

🚨 BREAKING: Microsoft launches a free Python library that converts ANY document to Markdown Introducing Markitdown. Let me explain. 🧵

29 226 2K 121K 2K

Download Image

Daily Dose of Data Science @DailyDoseOfDS_

2 weeks ago

The fastest serving engine for LLMs is here (open-source)! LMCache is an LLM serving engine designed to reduce time-to-first-token and increase throughput, especially under long-context scenarios. It boosts vLLM with 7x faster access to 100x more KV caches. 100% open-source!

16 207 1K 66K 1K

Download Image

Qwen @Alibaba_Qwen

3 weeks ago

🚀 Introducing Qwen3-Next-80B-A3B — the FUTURE of efficient LLMs is here! 🔹 80B params, but only 3B activated per token → 10x cheaper training, 10x faster inference than Qwen3-32B.(esp. @ 32K+ context!) 🔹Hybrid Architecture: Gated DeltaNet + Gated Attention → best of speed &…

172 723 4K 905K 2K

Download Image

Dan Mac @daniel_mac8

3 weeks ago

🚨 Did Moonshot just crack Continual Learning? Moonshot introduced: 'Checkpoint Engine'. It rapidly updates model weights in LLM inference engines. Can update a 1T param model like Kimi-K2 in 20 seconds. If this works, it truly is a *game changer* BIG. IF. TRUE. 🤯

37 126 1K 110K 905

Download Image

Sebastian Raschka @rasbt

3 weeks ago

Updated & turned my Big LLM Architecture Comparison article into a narrated video lecture. The 11 LLM architectures covered in this video: 1. DeepSeek V3/R1 2. OLMo 2 3. Gemma 3 4. Mistral Small 3.1 5. Llama 4 6. Qwen3 7. SmolLM3 8. Kimi 2 9. GPT-OSS 10. Grok 2.5 11. GLM-4.5

36 484 3K 166K 3K

Download Video

Tom Dörr @tom_doerr

3 weeks ago

AI health assistant that helps you manage and understand your health data

8 83 674 39K 854

Download Image

Qwen @Alibaba_Qwen

3 weeks ago

🎙️ Meet Qwen3-ASR — the all-in-one speech recognition model! ✅ High-accuracy EN/CN + 9 more languages: ar, de, en, es, fr, it, ja, ko, pt, ru, zh ✅ Auto language detection ✅ Songs? Raps? Voice with BGM? No problem. <8% WER ✅ Works in noise, low quality, far-field ✅ Custom…

82 337 2K 258K 949

Download Image

Akshay 🚀 @akshay_pachaar

4 weeks ago

Fine-tune 100+ LLMs directly from a UI! LLaMA-Factory lets you train and fine-tune open-source LLMs and VLMs without writing any code. Supports 100+ models, multimodal fine-tuning, PPO, DPO, experiment tracking, and much more! 100% open-source, 57k stars!