Feature/java bindings2 (#944)

whisper : add detect-language mode (#853)

whisper : add integer quantization support (#540)

whisper : fix typos in whisper.h (#737)

whisper : add progress callback (#600)

whisper : add initial_prompt param (#645)

main : provide option for creating JSON output (#615)

whisper.android : support benchmark for Android example. (#542)

whisper : add whisper_state + default state on the whisper_context (#523)

whisper : add API for applying custom logits filters during decoding

whisper : suppress non-speech-related token outputs (#473)

whisper : fixed Beam Search Strategy and exposed whisper_pcm_to_mel_phase_vocoder (#474)

whisper : add whisper_full_lang_id() for getting the context lang (#461)

whisper : add "split_on_word" flag when using using "max_len" option (#455)

bench : fix Windows linkage by moving ggml benches in whisper lib ..

stream : fix --keep_context argument to be used correctly (#354)

Improve decoding (#291)

ggml : remove obsolete zeroing + comment fixes (#390)

whisper : add loader class to allow loading from buffer and others (#353)

ggml : define MIN / MAX only if not defined (minor)

whisper : add whisper_n_audio_ctx and check for invalid audio_ctx

minor : resolves some of warnings when compiling with clang/clang++ (#294)

whisper : language auto-detect (#59)

whisper : add whisper_tokenize()

Try to improve the token sampling strategy (#193)

whisper : add mechanism for aborting the whisper_full() computation

whisper.objc : add real-time processing (#97)

whisper.objc : fix build warnings

whisper.wasm : do not block page while processing (close #86)

refactoring : more readable code

stream : "-kc" now enables context keeping from previous segment (#90)

Prompt previous tokens for streaming (#163)

stream : add "audio_ctx" parameter

stream : add "max_tokens" parameter

stream : add "single_segment" option

stream : partial encoder experiments

whisper : add option to speed up the audio tempo by x2

ref #22 : add "duration" option

whisper : token-level timestamp refactoring (#49, #120)

main : add option for word-leve timestamps (very experimental)

close #113 : fix struct whisper_token_data

parallel : print time of audio boundaries + fix timings

parallel : working

main : fix sampling time + add max_context parameter

parallel : adding tool for parallel transformer inference

ggml : add system info functions

whisper : add new-segment callback

wip : experimental color coding of tokens based on probabilities

ref #52 : improve greedy sampling strategy

close #32 : add comment about thread-safety of the C-style API