David Berenstein
@davidberen1957
Followers
8
Following
54
Media
2
Statuses
15
ML & DevRel @ Giskard & Pruna | ex HF ๐ค | ๐จ๐ฝโ๐ณ Cooking, ๐จ๐ฝโ๐ป Coding, ๐ Committing
Madrid
Joined July 2025
RT @Gradio: Video-Slider is a custom gradio component built by @pelolisu. pip install gradio_videoslider . More here ๐ .
huggingface.co
0
3
0
RT @MishaFein: Learn how our team cut ๐๐ ๐ข๐ง๐๐๐ซ๐๐ง๐๐ ๐ญ๐ข๐ฆ๐ ๐๐ฒ ๐๐% in just 2 days, without compromising quality. Some torch.compile + collaborโฆ.
blog.bria.ai
A real-world deep dive into optimizing GenAI inference latency with Torch Compile, LoRA hot-swapping, and Pruna AI.
0
1
0
๐ฑ EcoLogits tracks the energy consumption and environmental impacts of using generative AI models through APIs. Would you take a racecar to go grocery shopping? Some people would, but I prefer not to own one and take the bike ๐. GitHub repo:
0
0
1
RT @Bertrand_Charp: ๐โ๐บ ๐๐๐ฎ๐ฟ๐๐ถ๐ป๐ด ๐ฎ ๐ป๐ฒ๐ ๐๐ฒ๐ฟ๐ถ๐ฒ๐: ๐ฃ๐ฎ๐ฝ๐ฒ๐ฟ ๐๐ถ๐ด๐ต๐น๐ถ๐ด๐ต๐๐! Iโll share research papers about (efficient) AI Iโve read including theirโฆ.
0
2
0
RT @PrunaAI: ๐ฅJoin the Pruna x Salad webinar!. With @saladtech, we will show:.โข how to down and up scale deployment on consumer GPUs.โข howโฆ.
0
1
0
๐งโ๐ซ I gave a talk on Synthetic Data at PyCon Italy. Elevator pitch: "Got data problems? Relax. Synthetic data is here to help". If you want to learn how, make sure to check out the recording ๐. YouTube: .Slides:
0
0
0
RT @PrunaAI: ๐คฏย Qwen-Image just hits differently, especially if it is much cheaper ($0.025 per image) and faster!. A coffee shop entrance feโฆ.
0
1
0
๐ฅ We are handing out free red teaming! . If youโre building AI Agents, nowโs your chance to find the risks before attackers do. - Security review of your public AI bot.- Vulnerability detection for risks.- Actionable, expert guidance . Apply here:
giskard.ai
Prevent failures in your AI Agents with our free Red Teaming assessment. Giskard's Hub will detect vulnerabilities like prompt injections, data disclosure, sycophancy, and more.
0
0
0
RT @Bertrand_Charp: We are now close to real-time video generation: 5s video generated in 5s. Try it here:
replicate.com
The fastest Wan 2.2 text-to-image and image-to-video model
0
2
0
๐ Riding the wave of Wan 2.2 fast!. Only $0.05 and generated in roughly 20 seconds. Check the endpoint out: wan-video/wan-2.2-i2v-fast.
1
0
2
RT @replicate: We're pleased to work with Pruna to bring you a new and fast image model. It can generate 2 megapixel images in 3.4 secondsโฆ.
0
16
0
RT @PrunaAI: ๐ ๐ฃ๐ฟ๐๐ป๐ฎ ๐
@๐ด๐ผ๐ธ๐ผ๐๐ฒ๐ฏ ๐ฃ๐ฎ๐ฟ๐๐ป๐ฒ๐ฟ๐๐ต๐ถ๐ฝ ๐จ๐ฝ๐ฑ๐ฎ๐๐ฒ!. ๐ฅ Early adopters are reporting great results from our lightning-fast inference platforโฆ.
0
2
0