Speed up of AI-based pre-translations by parallelization of requests to the AI

XMLWordPrintable

    • Type: New Feature
    • Resolution: Unresolved
    • None
    • Affects Version/s: None
    • Component/s: translate5 AI

      Problem

      Current single threaded translation (sending only one batch of segments at a time) is to slow.

      Solution

      Speed up process as done already for quality estimation to send multiple batches of segments at the same time to the AI.

      Make it configurable, how many are send in parallel.

            Assignee:
            Aleksandar Mitrev
            Reporter:
            Marc Mittag [Administrator]
            Axel Becher, Leon Kiz
            Stephan Bergmann, Sylvia Schumacher
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

              Created:
              Updated: