Apple has publicly shared four open-source models boasting of enhanced accuracy for queries, which could help the development of future AI models. A Siri icon superimposed on Apple Park As the tech industry continues to race forward with AI developments, Apple has continued to offer more glimpses into the technology it is working on. In the latest public release, Apple has released a quartet of open-source models. Referred to as Open Source Efficient LLMs, or OpenELMs, the instruct...
Apple today released several open source large language models (LLMs) that are designed to run on-device rather than through cloud servers. Called OpenELM (Open-source Efficient Language Models), the LLMs are available on the Hugging Face Hub, a community for sharing AI code. As outlined in a white paper [PDF], there are eight total OpenELM models, four of which were pre-trained using the CoreNet library, and four instruction tuned models. Apple uses a layer-wise scaling strategy that is...
In a significant move towards enhancing privacy and processing efficiency, Apple has introduced a series of open source large language models (LLMs) known as OpenELM. These models are uniquely designed to operate directly on devices, diverging from the traditional reliance on cloud-based computations. This shift not only promises to improve user privacy by processing data […]
In terms of performance, the OpenLLM results shared by Apple show that the new models perform fairly well, especially the one with 3 billion parameters.
Ahead of iOS 18’s debut at WWDC in June, Apple has released a family of open-source large language models. Called OpenELM, Apple describes these as: a family of Open-source Efficient Language Models. In its testing, Apple says that OpenELM offers similar performance to other open language models, but with less training data. more
London-based model Alexsandrah has a twin, but not in the way you'd expect: Her counterpart is made of pixels instead of flesh and blood. The virtual twin was generated by artificial intelligence and has already appeared as a stand-in for the real-life Alexsandrah in a photo shoot. Alexsandrah, who goes by her first name professionally, in turn receives credit and compensation whenever the AI version of herself gets used - just like a human model. Alexsandrah says she and her alter-ego mirror...
Artificial intelligence is reshaping industries worldwide, and China is at the forefront of this technological revolution. The use of AI has been a hot topic at the China International Consumer Products Expo, where brands are exploring the potential
Enterprises still don't trust OpenAI, while open source AI models are cheaper and more than sufficient for targeted specific tasks, Nova AI founder says. Enterprises still don't trust OpenAI, while open source AI models are cheaper and more than sufficient for targeted specific tasks, Nova AI founder says.
aiXcoder 7B is a new open-source coding large language model with 7 billion parameters that has demonstrated superior performance in code generation tasks compared to other models like Code Llama 34B and Star Coder 15B. It supports multiple programming languages, including Python, JavaScript, C++, and others, and integrates with popular IDEs such as VS Code […]
Llama 3 is Meta’s latest generation of models that has state-of-the art performance and efficiency for openly available LLMs. Meta AI is available online for free. The small 7B model beats Mistral 7B and Gemma 7B. The 70B beats Claude 3 Sonnet (closed source Anthropic model) and competes against Gemini Pro 1.5 (closed source model
Zuckerberg says new AI model "was still learning" when Meta stopped training.
AI2 has released an update for its open source language model OLMo 1.7-7B with a new Dolma 1.7 dataset and two-stage training curriculum.