coreml : always convert to "neuralnetwork" (#2770)

whisper : add large-v3-turbo (#2440)

coreml : fix ANE optimized encoder (#1716)

whisper : make large version explicit + fix data size units (#1493)

coreml : use the correct `n_mel` value (#1458)

whisper : add support for large v3 (#1444)

coreml : add support of large-v1 model (#926)

minor : improve C++ and Python style (#768)

whisper : add Core ML support (#566)