Xenobd/whisper.cpp

1# whisper.cpp/examples/cli
2
3This is the main example demonstrating most of the functionality of the Whisper model.
4It can be used as a reference for using the `whisper.cpp` library in other projects.
5
6```
7./build/bin/whisper-cli -h
8
9usage: ./build/bin/whisper-cli [options] file0 file1 ...
10supported audio formats: flac, mp3, ogg, wav
11
12options:
13  -h,        --help              [default] show this help message and exit
14  -t N,      --threads N         [4      ] number of threads to use during computation
15  -p N,      --processors N      [1      ] number of processors to use during computation
16  -ot N,     --offset-t N        [0      ] time offset in milliseconds
17  -on N,     --offset-n N        [0      ] segment index offset
18  -d  N,     --duration N        [0      ] duration of audio to process in milliseconds
19  -mc N,     --max-context N     [-1     ] maximum number of text context tokens to store
20  -ml N,     --max-len N         [0      ] maximum segment length in characters
21  -sow,      --split-on-word     [false  ] split on word rather than on token
22  -bo N,     --best-of N         [5      ] number of best candidates to keep
23  -bs N,     --beam-size N       [5      ] beam size for beam search
24  -ac N,     --audio-ctx N       [0      ] audio context size (0 - all)
25  -wt N,     --word-thold N      [0.01   ] word timestamp probability threshold
26  -et N,     --entropy-thold N   [2.40   ] entropy threshold for decoder fail
27  -lpt N,    --logprob-thold N   [-1.00  ] log probability threshold for decoder fail
28  -nth N,    --no-speech-thold N [0.60   ] no speech threshold
29  -tp,       --temperature N     [0.00   ] The sampling temperature, between 0 and 1
30  -tpi,      --temperature-inc N [0.20   ] The increment of temperature, between 0 and 1
31  -debug,    --debug-mode        [false  ] enable debug mode (eg. dump log_mel)
32  -tr,       --translate         [false  ] translate from source language to english
33  -di,       --diarize           [false  ] stereo audio diarization
34  -tdrz,     --tinydiarize       [false  ] enable tinydiarize (requires a tdrz model)
35  -nf,       --no-fallback       [false  ] do not use temperature fallback while decoding
36  -otxt,     --output-txt        [false  ] output result in a text file
37  -ovtt,     --output-vtt        [false  ] output result in a vtt file
38  -osrt,     --output-srt        [false  ] output result in a srt file
39  -olrc,     --output-lrc        [false  ] output result in a lrc file
40  -owts,     --output-words      [false  ] output script for generating karaoke video
41  -fp,       --font-path         [/System/Library/Fonts/Supplemental/Courier New Bold.ttf] path to a monospace font for karaoke video
42  -ocsv,     --output-csv        [false  ] output result in a CSV file
43  -oj,       --output-json       [false  ] output result in a JSON file
44  -ojf,      --output-json-full  [false  ] include more information in the JSON file
45  -of FNAME, --output-file FNAME [       ] output file path (without file extension)
46  -np,       --no-prints         [false  ] do not print anything other than the results
47  -ps,       --print-special     [false  ] print special tokens
48  -pc,       --print-colors      [false  ] print colors
49  -pp,       --print-progress    [false  ] print progress
50  -nt,       --no-timestamps     [false  ] do not print timestamps
51  -l LANG,   --language LANG     [en     ] spoken language ('auto' for auto-detect)
52  -dl,       --detect-language   [false  ] exit after automatically detecting language
53             --prompt PROMPT     [       ] initial prompt (max n_text_ctx/2 tokens)
54  -m FNAME,  --model FNAME       [models/ggml-base.en.bin] model path
55  -f FNAME,  --file FNAME        [       ] input audio file path
56  -oved D,   --ov-e-device DNAME [CPU    ] the OpenVINO device used for encode inference
57  -dtw MODEL --dtw MODEL         [       ] compute token-level timestamps
58  -ls,       --log-score         [false  ] log best decoder scores of tokens
59  -ng,       --no-gpu            [false  ] disable GPU
60  -fa,       --flash-attn        [false  ] flash attention
61  -sns,      --suppress-nst      [false  ] suppress non-speech tokens
62  --suppress-regex REGEX         [       ] regular expression matching tokens to suppress
63  --grammar GRAMMAR              [       ] GBNF grammar to guide decoding
64  --grammar-rule RULE            [       ] top-level GBNF grammar rule name
65  --grammar-penalty N            [100.0  ] scales down logits of nongrammar tokens
66```
67