Advancements in hardware (H100 GPUs), software (CUDA, cuBLAS, cuDNN, FlashAttention), and data quality have drastically reduced training costs.
GAVEL essentially uses a a readily available dataset of games that were made using the Ludii game system, generating variations of the games that don’t exist.
“High-quality preference data is crucial for aligning AI systems with human values, but existing datasets are often proprietary or of inconsistent quality,” said Zhilin Wang, senior research scientist at NVIDIA.
Researchers believe that to make LLMs effectively understand or generate humor, their value-based alignment should be redirected from global alignment to community based alignment with specific audiences and comedians.
“Currently more than 50% of our employees are from India” said CEO Nayaki Nayyar.
Boasting 7 billion parameters, the model is built on Meta Llama-2+Mistral AI framework.
“One of the things that we are focused on is developing models that reach a vast audience, even in remote villages,” said Professor Maunendra Sankar Desarkar.
The first two KL3M models are kl3m-170m and kl3m-1.7b, designed for real-time use on consumer-grade hardware.
A self-taught AI enthusiast and developer, Vik Paruchuri, believes his OCR model, Surya, would help create low-resource Indic language datasets and models.
Amdocs’ partnerships with NVIDIA, and Microsoft Azure entails customisation of enterprise-grade AI framework for telcos.
Meta has unexpectedly become the Robinhood of the LLMbcommunity since its language model was leaked
In liquid neural networks, the parameters change over time based on the results of a nested set of differential equations, which means it understands new tasks by itself, and thus
Companies making generative models accessible are thriving more than a more impactful research
With these updates, Bard has become a genuine contender in the AI chatbot market, and looks to surpass competitors such as ChatGPT and Microsoft’s Bing Chat.
In an era when AI threatens to wipe out jobs, prompt engineering is an essential skill to learn to stay relevant.
ESP is building the technology on top of the research in the human domain, where they’re developing models that can work with bats and whales
The LLM revolution has made chips that can handle AI at a large scale more important than ever
“It surprised us all, including the people who are working on these things (LLMs). There’s been progressive improvement, but nobody really expected this level of human utility.”
The use of chatbots in healthcare is expected to grow due to ongoing investments in artificial intelligence and the benefits they provide
Every model before GPT-4 that researchers don’t trust the public with.
Every model open sourced since the genesis of GPT-2.
Ed Grefenstette spoke to AIM about his shift from big tech to startup.
Even though the enterprise was harnessing the powers of generative AI in 2022, if we talk about research, 2022 was definitely the year of protein fold predictions
“Understanding the theory requires a sophisticated understanding of physics”
If you want to learn more about the talk of the town — LLMs — you should definitely check out this list
ChatGPT’s parent, OpenAI has published a study to solve the problems of LLMs
Noam Chomsky believes that no matter how many models come with updated data and parameters, the fundamental flaw in the LLMs can never be remedied.
We are still in infancy, and to enable more companies to build solutions for mass consumption, it is important that we start building these open-source datasets: Raghu Dharmaraju
A new technique has enabled AI models to learn from data derived from devices. However, like many other models, it comes with its own set of challenges
The first version of Shoonya is expected to release later this month.
Tech mahindra news | Meta news | Semiconductor news | Mphasis news | Oracle news | Intel news | Deloitte news | Jio news | Job interview news | virtual internship news | IIT news | Certification news | Course news | Startup news | Leetcode news | claude news | Snowflake news | Python news | Microsoft news | AWS news
Discover how Cypher 2024 expands to the USA, bridging AI innovation gaps and tackling the challenges of enterprise AI adoption
© Analytics India Magazine Pvt Ltd & AIM Media House LLC 2024
