• Tuesday • July 16, 2024

    Avatar photo

    Industry News Source

    Hugging Face’s SmolLM models bring powerful AI to your phone, no cloud required

    These Models Bring Advanced AI Capabilities to Personal Devices

    Hugging Face today unveiled SmolLM, a new family of compact language models that surpass similar offerings from Microsoft, Meta, and Alibaba’s Qwen in performance. The SmolLM lineup features three sizes — 135 million, 360 million, and 1.7 billion parameters — designed to accommodate various computational resources. Despite their small footprint, these models have demonstrated superior results on benchmarks testing common sense reasoning and world knowledge.

    Loubna Ben Allal, lead ML engineer on SmolLM at Hugging Face, emphasized the efficacy of targeted, compact models in an interview with VentureBeat. “We don’t need big foundational models for every task, just like we don’t need a wrecking ball to drill a hole in a wall,” she said. “Small models designed for specific tasks can accomplish a lot.”

    ,
  • Friday • July 12, 2024

    Avatar photo

    Industry News Source

    Standard Bots reels in $63M for its AI-powered robotic arms

    The VC Firm Was Joined by the Amazon Industrial Innovation Fund and Samsung Next

    Standard Bot’s flagship product is a $37,000 robotic arm called the RO1. It can lift up to 40 pounds and has a reach of about four feet. The RO1 comes with a portable base, which removes the need to embed the robot into a factory floor’s concrete floor and thereby reduces installation costs.

    Under the hood, the system ships with a built-in graphics processing unit. The RO1 uses the chip to run artificial intelligence models that allow it to complete tasks even in the face of unexpected challenges. The algorithms can maneuver the robotic arm around any objects that might be in its way, as well as adapt to production line changes.

    ,
  • Thursday • July 11, 2024

    Avatar photo

    Industry News Source

    Fireworks AI Valued at $552 Million After New Funding Round

    Qiao Said The Industry is Heading to a Place Where Using a Mix of AI Systems Becomes the Norm

    “Unlike proprietary mega models that are generic, non-private, and hard to customize, Fireworks AI provides smaller, production-grade models that can be deployed privately and securely,” CEO Lin Qiao wrote in the post. “Using minimal human-curated data, our ultra-fast LoRA fine-tuning allows developers to quickly customize models to their specific needs, transitioning from dataset preparation to querying a fine-tuned model in minutes. These fine-tuned models are seamlessly deployed, maintaining the same performance and cost benefits as our base models.”

    Developers at AI startups such as Cresta, Cursor and Liner have used Fireworks’ AI models, as have tech companies including DoorDash, Quora and Upwork, the post said.

    ,
  • Wednesday • June 26, 2024

    Avatar photo

    Techcrunch

    Formation Bio raises $372M to boost drug development with AI

    It’s a “Material Step Up” From $1 Billion, Formation’s Series Valuation, said a Spokesperson

    Formation announced Wednesday that it raised $372 million in a Series D funding round led by Andreessen Horowitz with participation from drug maker Sanofi, Sequoia, Thrive, Emerson Collective, Lachy Groom, SV Angel Growth and FPV Ventures. The new tranche brings Formation’s total raised to more than $600 million (according to PitchBook), which the company says is being put mainly toward partnership acquisition efforts and R&D.

    Formation builds tech-forward solutions for clinical trials and drug development. The company licenses drug IP from and co-develops drugs with biotech and pharma companies, and develops these drugs past clinical proof-of-concept.

    Drug development is a notoriously expensive and challenging endeavor. It takes 10 to 15 years on average to take a drug from initial discovery through regulatory approval, with the cost per drug reaching up to $5.5 billion. And an estimated 90% of drugs fail to reach the market.

    , ,
  • Friday • June 21, 2024

    Avatar photo

    Industry News Source

    Anthropic’s Claude 3.5 Sonnet beats GPT-4o in most benchmarks

    The Model Demonstrates Enhanced Capabilities in Understanding Nuance, Humor, and Complex Instructions

    Claude 3.5 Sonnet is now accessible for free on Claude.ai and the Claude iOS app, with higher rate limits for Claude Pro and Team plan subscribers. It’s also available through the Anthropic API, Amazon Bedrock, and Google Cloud’s Vertex AI. The model is priced at $3 per million input tokens and $15 per million output tokens, featuring a 200K token context window.

    Anthropic claims that Claude 3.5 Sonnet “sets new industry benchmarks for graduate-level reasoning (GPQA), undergraduate-level knowledge (MMLU), and coding proficiency (HumanEval).” The model demonstrates enhanced capabilities in understanding nuance, humour, and complex instructions, while excelling at producing high-quality content with a natural tone.

    Operating at twice the speed of Claude 3 Opus, Claude 3.5 Sonnet is well-suited for complex tasks such as context-sensitive customer support and multi-step workflow orchestration. In an internal agentic coding evaluation, it solved 64% of problems, significantly outperforming Claude 3 Opus at 38%.

    ,
  • Thursday • June 20, 2024

    Avatar photo

    Techcrunch

    Daydream rakes in $50M seed funding to build an AI-powered search engine suited for e-commerce

    Daydream Focuses on Providing Personalized Shopping Results by Using Generative AI

    The company plans to offer up the platform in beta to consumers in the U.S. this fall, and will concentrate on the fashion vertical for the time being. It has already onboarded more than 2,000 brands, including Net-A-Porter, Altuzarra, Jimmy Choo, Doen, Alo Yoga and La DoubleJ.

    Daydream essentially lets one search for products using natural language and image recognition — think queries like “I’m going to a wedding in Costa Rica in the summer, and I need some suggestions of what to wear.” You can even upload an image of a piece of clothing and provide an additional filter with conversational language, such as “I want this in blue.”

    Forerunner Ventures and Index Ventures co-led the round, which also saw participation from Google Ventures and True Ventures. The startup did not disclose its valuation.

    , , ,