Загрузка видео...

Не удалось загрузить видео

На главную

Google released MedGemma on I/O'25 👏 > 4B and 27B instruction fine-tuned vision LMs and a 4B pre-trained vision LM for medicine > available with transformers from the get-go 🤗 they also released a cool demo for scan reading ⤵️

81,353 просмотров • 1 год назад •via X (Twitter)

Комментарии: 8

Фото профиля merve
merve1 год назад

demo: models: use with transformers ⤵️

Фото профиля Sterling Cooley
Sterling Cooley2 лет назад

Hey! I wanted to make sure you saw we're doing a Live Webinar for the Ultra Skool Learn how to use Ultrasound Vagus Nerve Stimulation - people are having absolutely WILD experiences on this

Фото профиля Aylin Gümüş
Aylin Gümüş1 год назад

i tried the demo, it looks so cool! it'd be such a huge and meaningful contribution to the health industry! 🦾

Фото профиля C
C1 год назад

@huggingface within minutes, health care giants file patents for new image formats to make this impossible and lobby lawmakers to make health data proprietary and keep this sort of thing from ever becoming useful. and they've already done it.

Фото профиля poyraz
poyraz1 год назад

Has Gemma Med been trained on abdominal CT scans?

Фото профиля Roberto Tomás C 🍉
Roberto Tomás C 🍉1 год назад

Didn’t they just release a txGemma too? What’s the difference?

Фото профиля Vishal
Vishal1 год назад

Huge step for AI in healthcare! Excited to see how MedGemma helps with medical imaging and makes expert tools more accessible.

Фото профиля Sam Woods
Sam Woods1 год назад

Great to see Google pushing more specialized AI for healthcare

Похожие видео

Chinese robotics company Astribot released their latest World-Action Model (WAM), Lumo-2. Technical breakdown: - based on a frozen 🥶 Qwen-3.5 4B VLM - trained in 3 progressive stages: 1. Action is aligned with latent world dynamics (an abstract representation of action). Real-world actions are anchored to physical constraints, while the latent space is guided to focus on motion-relevant changes. This bidirectional relationship makes the model physically grounded -> critical for a world model. 2. Action is aligned with vision and language. Reusing the vision backbone and action encoder from the frozen VLM, the authors add a custom vocabulary (for new actions), a semantic module, an action decoder, and an action projector. This aligns the (new) action representations with the (existing) vision-language semantic space. Most importantly: it builds a direct mapping from natural-language instructions to motor execution. 3. End-to-end training on language, video, and robot data. Only the new modules (everything outside the frozen backbone) are trained end-to-end across temporal reasoning, physical understanding, long-horizon, and dexterous manipulation. At the end of the day, Lumo-2 is not the best on benchmarks, but that's not the point. What's genuinely new: - a way to combine latent world modeling and action generation through progressive alignment - a physically-grounded latent dynamics space - it lifts performance on unseen objects using un-annotated human egocentric video + Vision Pro captures, no special transfer algorithm needed Why it matters: - the whole model is thin trainable adapters (semantic module, action decoder/projector) on a frozen 4B backbone (cheap) - that scale is suited for real-time embedded inference (~2.71× decode speedup, no accuracy loss) - its real moat is long-horizon execution, where the added temporal memory pays off far more than on any other task As a result, this robot can now make your latte (5x sped up video):

Léo

32,296 просмотров • 1 месяц назад

Molmo by Ai2 - Open source SoTA Multimodal (Vision) Language model, beating Claude 3.5 Sonnet, GPT4V and comparable to GPT4o 🔥 They release four model checkpoints: 1. MolmoE-1B, a mixture of experts model with 1B (active) 7B (total) 2. Molmo-7B-O, most open 7B model 3. Molmo-7B-D, demo model 4. Molmo-72B, best model System Architecture > Input: Multi-scale, multi-crop images generated from the original image. > Vision Encoder: OpenAI's ViT-L/14 336px CLIP model, a powerful ViT, encodes images into vision tokens. > Connector: MLP projects tokens to LLM input space, followed by pooling for dimensionality reduction. > LLM: Decoder-only Transformer, various options (OLMo, OLMoE, Qwen2, Mistral, Gemma2, Phi) with diverse scales and openness. Model Variants > Vision Encoder: Consistent ViT-L/14 CLIP model across variants. > LLM: OLMo-7B-1024, OLMoE-1B-7B-0924, Qwen2 (7B, 72B), Mistral 7B, Gemma2 9B, Phi 3 Medium, offering different capacities and openness levels. Training Strategy > Stage 1: Multimodal pre-training for caption generation with new captioning data. > Stage 2: Supervised fine-tuning on a dataset mixture, updating all parameters. > No RLHF involved, Learning rates adjusted based on component types and pre-training status. > All the weights are available on Hugging Face Hub 🤗 > Compatible with Transformers (Remote Code) Kudos Ai2 for such a brilliant and open work! 🐐 Video credits: Allen AI YT Channel

Vaibhav (VB) Srivastav

80,474 просмотров • 1 год назад