evaluate the decoder given text prompt + audio features -> predicts the probabilities for the next token - model: the model - n_threads: numbe

evaluate the encoder given audio recording (more specifically, its log mel spectrogram), runs forward pass of the encoder part of the transformer mod

the most basic sampling scheme - select the top token

wrap the last segment to max_len characters returns the number of new segments

Common Control Versions: Win95/WinNT 4.0 maj=4 min=00 IE 3.x maj=4 min=70 IE 4.0 maj=4 min=71

Create methods

for listbox items

single-selection only

Operations

Note!: This function leaves the stream in an unclean/broken state! * It is not safe to subsequently use the same state with a _fastReset() or * _con

LZ4_compress_generic() : * inlined, to ensure branches are decided at compilation time. * Presumed already validated at this stage: * - source !

The "double dictionary" mode, for use with e.g. ring buffers: the first part * of the dictionary is passed as prefix, and the second via dictStart +

-****************************** * Compression functions ********************************/

LZ4_memcpy_using_offset() presumes : * - dstEnd >= dstPtr + MINMATCH * - there is at least 8 bytes available to write after dstEnd */

Create/destroy methods

Overrideables