Tech giant Meta is making it's much talked about language model LLaMA2 free for all.
Facebook’s parent firm announced the open sourcing of LLaMA2 for the sake of commercial as well as research purposes. And the news means it’s competing head to head against OpenAI’s GPT-4 which also works free for all. The latter is used to power popular tools in today’s use such as the Bing chatbot by Microsoft and OpenAI’s ChatGPT.
The company mentioned today how the decision was being taken as a part of the trending Inspire event that’s being hosted by Microsoft and it was also going to provide support for platforms like Windows and Azure too. Similarly, it hoped that this would signal the start of a great collaborative effort with plenty of learning opportunities along the way.
Computational giant Microsoft mentioned more in detail how so many of its leading AI tools were getting incorporated into the 360 platforms and also what the estimated costs linked to the endeavor would be.
In the same way, we saw Qualcomm delineate how it would be busy working by the side of Meta to roll out laptops with the language model embedded into them. The same would be the case for smartphones and headsets.
We should be seeing an expected launch to arise as early as next year and continue from there onwards as it’s designed for apps that are powered by breakthrough AI technology and can function easily without reliance on things like Cloud Services.
Meanwhile, a recently published press release went on to explain how this whole initiative was designed to allow firms including startups to gain access to more tools of AI and enable members of the community to carry out trials and experiments with them and get familiar along the way.
As per statements made by Meta recently, the language model did receive training on around 40% more types of information when you compare it to the language model LLaMa 1 which entails data from various sources seen online. The latter also outperforms other language models such as MPT in terms of coding, knowledge, and also proficiency.
While putting out LLaMA in the open, Meta claims it wishes to enhance safety as well as transparency in this matter. It has already tested the model for things like privacy and safety via adversarial prompts.
The company also delineated how such models are tested and how changes are made to them.
The new LLaMa2 is going to be rolled out via Microsoft’s famous Azure initiative and a few other retailer providers.
The tech giant similarly mentioned how taking on such an approach was the right way to move ahead and also key for the growth of AI models including those related to the world of Generative AI where data and technology are at an all-time high.
So by opening them up to all and giving access to anyone and everyone out there, we’re seeing a greater chance of more progress and how more problems can be solved quicker for the entire community.
Read next: AI Research Wars: Meta's Open Innovation Showdown!
Facebook’s parent firm announced the open sourcing of LLaMA2 for the sake of commercial as well as research purposes. And the news means it’s competing head to head against OpenAI’s GPT-4 which also works free for all. The latter is used to power popular tools in today’s use such as the Bing chatbot by Microsoft and OpenAI’s ChatGPT.
The company mentioned today how the decision was being taken as a part of the trending Inspire event that’s being hosted by Microsoft and it was also going to provide support for platforms like Windows and Azure too. Similarly, it hoped that this would signal the start of a great collaborative effort with plenty of learning opportunities along the way.
Computational giant Microsoft mentioned more in detail how so many of its leading AI tools were getting incorporated into the 360 platforms and also what the estimated costs linked to the endeavor would be.
In the same way, we saw Qualcomm delineate how it would be busy working by the side of Meta to roll out laptops with the language model embedded into them. The same would be the case for smartphones and headsets.
We should be seeing an expected launch to arise as early as next year and continue from there onwards as it’s designed for apps that are powered by breakthrough AI technology and can function easily without reliance on things like Cloud Services.
Meanwhile, a recently published press release went on to explain how this whole initiative was designed to allow firms including startups to gain access to more tools of AI and enable members of the community to carry out trials and experiments with them and get familiar along the way.
As per statements made by Meta recently, the language model did receive training on around 40% more types of information when you compare it to the language model LLaMa 1 which entails data from various sources seen online. The latter also outperforms other language models such as MPT in terms of coding, knowledge, and also proficiency.
While putting out LLaMA in the open, Meta claims it wishes to enhance safety as well as transparency in this matter. It has already tested the model for things like privacy and safety via adversarial prompts.
The company also delineated how such models are tested and how changes are made to them.
The new LLaMa2 is going to be rolled out via Microsoft’s famous Azure initiative and a few other retailer providers.
The tech giant similarly mentioned how taking on such an approach was the right way to move ahead and also key for the growth of AI models including those related to the world of Generative AI where data and technology are at an all-time high.
So by opening them up to all and giving access to anyone and everyone out there, we’re seeing a greater chance of more progress and how more problems can be solved quicker for the entire community.
Read next: AI Research Wars: Meta's Open Innovation Showdown!
