🚀 Medial Secures Investment on Shark Tank India - Fueling the Future of Professional Social Networking. 🔥

Startup Showcase

Premium Content

Try our Valuation Calculator →

News on Medial

Meta releases Llama 3, claims it's among the best open models available

TechCrunch · 1y ago

Meta releases Llama 3, claims it's among the best open models available

Meta has open sourced two models in its Llama 3 series of generative AI models, with the remaining models to follow at a later date. The models, Llama 3 8B and Llama 3 70B, are said to be a significant improvement over the previous Llama models, both in terms of performance and number of parameters. Llama 3 8B outranks other open source models on several benchmarks, while Llama 3 70B competes with flagship models. The Llama 3 models boast a larger dataset and claim to provide improved steerability and accuracy. Meta plans to release multilingual and multimodal versions in the future.

Related News

Meta releases ChatGPT-like AI site and open-weights Llama 3 model

Arstechnica

Arstechnica · 1y ago

Meta releases ChatGPT-like AI site and open-weights Llama 3 model

Meta has introduced the Llama 3 AI model, which can be utilized for tasks like text composition, code generation, and chatbots. Available in 8 billion and 70 billion parameter sizes, Llama 3 is not classified as open-source but as open-weights. Meta's CEO, Mark Zuckerberg, mentioned training the 70 billion model on around 15 trillion tokens of data. Additionally, Meta announced plans to train a massive 400 billion parameter version of Llama 3. The company also disclosed that Llama 3 will be incorporated into the Meta AI assistant, accessible through major social media apps and websites.

Meta releases new AI model Llama 4

Economic Times

Economic Times · 4m ago

Meta releases new AI model Llama 4

Meta Platforms has unveiled Llama 4, its newest large language model, available in variants Llama 4 Scout and Llama 4 Maverick. These multimodal AI systems can process and integrate various data types like text, video, and audio. Llama 4's open-source nature aims to offer the best in multimodality. Additionally, Llama 4 Behemoth, a highly advanced teacher model, is being previewed. Meta is investing significantly in AI infrastructure, driven by competition and market expectations.

LLMs keep leaping with Llama 3, Meta’s newest open-weights AI model

Arstechnica

Arstechnica · 1y ago

LLMs keep leaping with Llama 3, Meta’s newest open-weights AI model

Meta has introduced Llama 3, an open-weights AI model for text composition, code generation, and chatbots. It is available in two versions, with parameter sizes of 8 billion and 70 billion. Meta is also training a 400 billion parameter version. Llama 3 performs well in benchmarks, such as undergraduate knowledge, grade-school math, coding, and graduate-level questions. The model has been trained with both images and text, with plans to support multimodal outputs in the future. Meta's AI assistant, powered by Llama 3, will be integrated into major social media apps and search features.

Meta’s free Code Llama AI programming tool closes the gap with GPT-4

The Verge

The Verge · 1y ago

Meta’s free Code Llama AI programming tool closes the gap with GPT-4

Meta has released an update to its Code Llama 70B AI model, claiming it to be their best-performing model yet. Code Llama 70B has improved accuracy compared to previous versions, allowing developers to provide more prompts while programming. Scoring better than GPT-3.5, this model achieved a 53% accuracy on the HumanEval benchmark. Code Llama helps developers create and debug code, and it is available for research and commercial use on Hugging Face. Meta believes larger models like Llama 70B provide better coding assistance.

Meta unveils biggest Llama 3 AI model, touting language and math gains

Economic Times

Economic Times · 1y ago

Meta unveils biggest Llama 3 AI model, touting language and math gains

Meta Platforms has released Llama 3, an AI model with 405 billion parameters that can converse in eight languages, write high-quality code, and solve complex math problems. The new model outperforms previous versions and rivals some paid models from competitors like OpenAI. The release of Llama 3 aligns with Meta's strategy to offer free models to developers, with CEO Mark Zuckerberg expecting future Llama models to surpass proprietary competitors by next year. Meta also plans to release multimodal versions of Llama 3 later this year, incorporating image, video, and speech capabilities.

Meta permits its AI models to be used for US military purposes

Economic Times

Economic Times · 9m ago

Meta permits its AI models to be used for US military purposes

Meta is now allowing US government agencies and contractors to use its AI models for military purposes, a departure from its previous policy. The company's AI models, called Llama, will be made available to federal agencies and already has partnerships with defense contractors like Lockheed Martin and Booz Allen. Llama is open source, meaning it can be freely copied and distributed. Meta's move aims to support the safety, security, and economic prosperity of the United States and its allies in the global AI race. The company plans to share its technology with members of the Five Eyes intelligence alliance.

Meta releases two Llama 4 AI models

The Verge

The Verge · 2m ago

Meta releases two Llama 4 AI models

Meta has unveiled its Llama 4 AI models, comprising Llama 4 Scout and Llama 4 Maverick, claiming superior performance over counterparts from OpenAI and Google across numerous benchmarks. Llama 4 Scout is noted for its efficiency, operating within a single Nvidia H100 GPU. Meanwhile, Llama 4 Behemoth, currently in training, boasts advanced parameters expected to outperform competitors. Meta's "mixture of experts" architecture optimizes resource use, though licensing restrictions challenge its open-source label.

In AI copyright case, Zuckerberg turns to YouTube for his defense | TechCrunch

TechCrunch · 6m ago

In AI copyright case, Zuckerberg turns to YouTube for his defense | TechCrunch

Meta CEO Mark Zuckerberg is using YouTube's approach to handling pirated content as a defense in an AI copyright case involving Meta. The case is centered around the use of a dataset containing copyrighted e-books for training Meta's AI models, Llama. Zuckerberg argues that, similar to YouTube, it's reasonable to use such datasets despite potential copyright issues. Plaintiffs, including authors like Sarah Silverman, claim Meta cross-referenced pirated books with copyrighted ones, seeking potential licensing.

Meta AI Expands Language Support With Hindi And Hindi-Romanised Script

Inc42 · 1y ago

Meta AI Expands Language Support With Hindi And Hindi-Romanised Script

Meta AI, the artificial intelligence platform, has expanded its language support to include seven new languages, including Hindi. The company is now available in 22 countries, with recent additions such as Argentina, Chile, Colombia, Ecuador, Mexico, Peru, and Cameroon. Additionally, Meta AI has released its advanced open-source model, Llama 405B, which is now accessible on WhatsApp and meta.ai in Hindi.

Zuckerberg’s AGI remarks follow trend of downplaying AI dangers

Arstechnica

Arstechnica · 1y ago

Zuckerberg’s AGI remarks follow trend of downplaying AI dangers

Meta CEO Mark Zuckerberg announced that the company is working on developing "general intelligence" for AI assistants and plans to open source it responsibly. This move involves bringing together Meta's research groups FAIR and GenAI. While Zuckerberg did not explicitly mention "artificial general intelligence" (AGI), the announcement suggests a focus on that direction. However, Zuckerberg downplayed the immediate threat of AGI and emphasized the gradual development of intelligence over time. This contrasts with the previous hype surrounding AGI. The announcement also mentioned the training of Llama 3 and the massive GPU capacity that Meta is amassing for AI models.

Trackers

Active Indian VC’s

OG Capital Email

With a hands-on approach, OG Capital aims to invest in over 20 promising...

Accel Partners Email

Early and growth-stage investments in disruptive technology companies with...

Early-stage venture capital firm investing in technology startups in India. Focus on...

Access All Trackers

Startup Showcase Winners

June 2025

Helping your parents when you are miles away

The Pit Stop Your Cravings Deserve

The next generation E-commerce platform

Enter Ongoing Startup Showcase

Top Users

Trending News on Medial

Download the medial app to read full posts, comements and news.