xAI, the artificial intelligence startup owned by Elon Musk, today launched an upgraded version of its chatbot Grok, called Grok-1.5.
The company says the new version will soon be available to early testers and current Grok users on social media platform X (formerly Twitter).
“One of the most significant improvements to Grok-1.5 is its performance in coding and math-related tasks,” xAI said in a statement.
Grok and Grok 1.5
Performance
-Grok achieved a score of 23.9% on the MATH benchmark and 62.9% on the GSM8K benchmark.
-Grok 1.5 shows significant improvements, scoring 50.6% on the MATH benchmark and 90% on the GSM8K benchmark. It also scored 74.1% on the HumanEval benchmark, indicating improved code generation and problem-solving capabilities.
Understand long context
-Grok has limited ability to handle long contexts.
-Grok 1.5 can now understand and process longer text fragments, up to 128,000 tokens. This enables it to efficiently find and use information in larger documents, making it more useful when retrieving specific details from longer texts.
infrastructure
-Grok runs on existing infrastructure.
-Grok 1.5 is built on a custom distributed training framework based on JAX, Rust and Kubernetes. This infrastructure allows for the flexibility to train large language models on large-scale GPU (graphics processing unit) clusters, which helps ensure that the training process is robust and reliable, reducing the possibility of problems or delays.
Availability
-Grok is currently available for use.
-Grok 1.5 will be available to early testers soon, with wider rollout planned in the near future. xAI has urged users to provide feedback to help further improve the platform.
Follow us on Google news ,Twitter , and Join Whatsapp Group of thelocalreport.in