This dataset, collected from various LLM inference services on Azure on November 11th, 2023, serves as the basis for the data described and analyzed in the ISCA 2024 paper titled 'Splitwise: Efficient generative LLM inference using phase splitting'.
Sreebhargavibalijaa/Microsoft-Azure-LLM-s-inference
Performing analysis on Microsoft azure LLM inference rates
Jupyter Notebook