St. LEO V: Speech -> Text on LEO5

About this dataset This project was created to provide the Faculty of Social and Political Sciences at the University of Innsbruck with a tool for transcribing audio-recorded interviews and attributing speakers, as well as enabling automatic translation if required. Purpose This dataset documents version 1.1 of the corresponding GitLab repository, which was used to develop and distribute the tool. Since it is general enough to be useful to other operators of high-performance clusters who want to provide a similar setup, it is made available here. Creation This is done by generating a tagged v1.1 in the GitLab repository https://git.uibk.ac.at/zid/hpc/st-leo-v/-/tree/072190574b8e96be514d85a20dabe1923d440377/ Technical details README.md directly mirrors v1.1 of the gitlab readme, the archive contains the corresponding repo (for decompression please resort to standard tools for your operating system). Funding This work got support from EuroCC Austria, the Austrian National Competence Centre for High-Performance Computing, High-Performance Data Analytics and Artificial Intelligence, which has received funding from the European High Performance Computing Joint Undertaking (JU) and Germany, Bulgaria, Austria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, Greece, Hungary, Ireland, Italy, Lithuania, Latvia, Poland, Portugal, Romania, Slovenia, Spain, Sweden, France, Netherlands, Belgium, Luxembourg, Slovakia, Norway, Türkiye, Republic of North Macedonia, Iceland, Montenegro, Serbia under grant agreement No 101101903.