Ethan B. Holland

ModelSpace/GemmaX2-28-2B-v0.1 · Hugging Face https://huggingface.co/ModelSpace/GemmaX2-28-2B-v0.1

“@RitvikRastogi19 yo yo, they released their tech report here: https://x.com/reach_vb/status/1894991762202259530

“Just landed FlashMLA in vLLM and it is already boosting output throughput 2-16%! We expect more improvements in the coming days as we continue to optimize the code path. https://x.com/vllm_project/status/1894994674630435123

“Microsoft released Phi-4-multimodal (5.6B parameters) and Phi-4-mini (3.8B) to consolidate speech, vision, and text in one efficient model family, boosting low-latency deployment on resource-limited environments. 🔑 Key Features Phi-4-multimodal handles audio, images, and text https://x.com/rohanpaul_ai/status/1894900923430064291

“Announcing Minions: a method that pairs small language models on a laptop (@ollama) with frontier models in the cloud—preserving 98% of the accuracy for <18% of the cost! Led by @Avanika15, @EyubogluSabri, and @dan_biderman (@HazyResearch), with @togethercompute’s @avnermay. 🧵 https://x.com/togethercompute/status/1894392054043578373

“Github 👨‍🔧: AI app store powered by 24/7 desktop history. open source | 100% local | dev friendly | 24/7 screen, mic recording ————- → 24/7 local screen and microphone recording. → API access to indexed desktop activity history. → Plugin system (“pipes”) for https://x.com/rohanpaul_ai/status/1893075129254699327

“Stagehand v1.13 has been released! 1. Customize local browser with cookies, viewport, and more 2. Act() from a stale observe result will automatically attempt to self-heal 3. Better developer experience around iframes 4. Novel DOM-a11y hybrid data structure Details in 🧵” / X https://x.com/Stagehanddev/status/1892644761338585595

“Boris, check out our mini model numbers, it surpassed o3mini high in all AIME 2024, GPQA, and LCB for pass@1. Generally I also don’t think our current benchmarks capture enough of the model intelligence. Our big Grok3 is worse on pass@1, but in our testing we can feel a smarter https://x.com/Yuhu_ai_/status/1892449337218883868

“Minions: Cost-efficient Collaboration Between On-device and Cloud Language Models “MinionS, in which the remote model decomposes the task into easier subtasks over shorter chunks of the document, that are executed locally in parallel. MinionS reduces costs by 5.7× on average https://x.com/iScienceLuvr/status/1894354075757777311

“Small Models Struggle to Learn from Strong Reasoners https://x.com/_akhaliq/status/1892435858684248326

“Phi 4 Mini Text: https://x.com/reach_vb/status/1894992084223889563

Minions: the rise of small, on-device LMs · Hazy Research https://hazyresearch.stanford.edu/blog/2025-02-24-minions

“Microsoft just dropped Phi-4 models on @huggingface 🧠 Phi-4-multimodal (5.6B) tops ASR leaderboards with 6.14% word error rate 👁️ Strong vision capabilities for math & science reasoning 📊 Phi-4-mini (3.8B) outperforms larger models in reasoning tasks 📏 Supports up to 128K https://x.com/fdaudens/status/1894884795856900505

“Miquel and team COOKED: SmolVLM2 – Apache 2.0 licensed VideoLMs (ranging from 2.2B to 256M) – can even run on a FREE colab🔥 Beats models 5x it’s size – runs on an iPhone! Trained with memory efficiency as a focus you can pass extremely long videos with little VRAM required🤗 https://x.com/reach_vb/status/1892578169615523909

“Check-out SmolVLM2 with day-zero support for MLX + MLX Swift to run locally on any Apple device. Small yet powerful enough to do fast *video* understanding on an iPhone! https://x.com/awnihannun/status/1892594913893556707

“Introducing Perplexity’s new voice mode. Ask any question. Hear real-time answers. Update your iOS app to start using. Coming soon to Android and Mac app. https://x.com/perplexity_ai/status/1894788583770509505

“we just dropped SmolVLM2: world’s smollest video models in 256M, 500M and 2.2B ⏯️🤗 we also release the following 🔥 > an iPhone app (runs on 500M model in MLX) > integration with VLC for segmentation of descriptions (2.2B) > a highlights extractor (2.2B) https://x.com/mervenoyann/status/1892576290181382153

Empowering innovation: The next generation of the Phi family | Microsoft Azure Blog https://azure.microsoft.com/en-us/blog/empowering-innovation-the-next-generation-of-the-phi-family/

“Phi 4 Multimodal Instruct: https://x.com/reach_vb/status/1894991935456124941

“Turns out this is powered by a 7B parameter onboard vision based LLM… ok this is interesting and sort of expected (had this exact discussion in 2023), difference is now the models are more capable” / X https://x.com/abacaj/status/1892611093802910152

SmolVLM2: Bringing Video Understanding to Every Device https://huggingface.co/blog/smolvlm2