mirror of
https://github.com/ggerganov/whisper.cpp.git
synced 2025-07-04 08:20:57 +02:00
Compare commits
1 Commits
| Author | SHA1 | Date | |
|---|---|---|---|
| 7aa1174315 |
5
.github/workflows/bindings.yml
vendored
5
.github/workflows/bindings.yml
vendored
@ -3,11 +3,6 @@ on:
|
|||||||
push:
|
push:
|
||||||
paths:
|
paths:
|
||||||
- bindings/go/**
|
- bindings/go/**
|
||||||
- whisper.h
|
|
||||||
pull_request:
|
|
||||||
paths:
|
|
||||||
- bindings/go/**
|
|
||||||
- whisper.h
|
|
||||||
|
|
||||||
jobs:
|
jobs:
|
||||||
ubuntu-latest:
|
ubuntu-latest:
|
||||||
|
|||||||
2
.github/workflows/build.yml
vendored
2
.github/workflows/build.yml
vendored
@ -1,5 +1,5 @@
|
|||||||
name: CI
|
name: CI
|
||||||
on: [push, pull_request]
|
on: [push]
|
||||||
|
|
||||||
jobs:
|
jobs:
|
||||||
ubuntu-latest:
|
ubuntu-latest:
|
||||||
|
|||||||
48
.github/workflows/examples.yml
vendored
48
.github/workflows/examples.yml
vendored
@ -1,48 +0,0 @@
|
|||||||
name: Examples Tests
|
|
||||||
on:
|
|
||||||
push:
|
|
||||||
paths:
|
|
||||||
- examples/addon.node/**
|
|
||||||
- whisper.h
|
|
||||||
pull_request:
|
|
||||||
paths:
|
|
||||||
- examples/addon.node/**
|
|
||||||
- whisper.h
|
|
||||||
|
|
||||||
jobs:
|
|
||||||
addon_node-ubuntu-latest:
|
|
||||||
runs-on: ubuntu-latest
|
|
||||||
strategy:
|
|
||||||
matrix:
|
|
||||||
node-version: [ 16.x, 18.x ]
|
|
||||||
steps:
|
|
||||||
- name: Clone
|
|
||||||
uses: actions/checkout@v1
|
|
||||||
|
|
||||||
- name: Dependencies
|
|
||||||
run: |
|
|
||||||
sudo apt-get update
|
|
||||||
sudo apt-get install build-essential
|
|
||||||
sudo apt-get install cmake
|
|
||||||
sudo apt-get install libsdl2-dev
|
|
||||||
|
|
||||||
- name: Use Node.js ${{ matrix.node-version }}
|
|
||||||
uses: actions/setup-node@v1
|
|
||||||
with:
|
|
||||||
node-version: ${{ matrix.node-version }}
|
|
||||||
cache: 'npm'
|
|
||||||
|
|
||||||
- name: Install package.json dependencies
|
|
||||||
working-directory: ./examples/addon.node
|
|
||||||
run: npm install
|
|
||||||
|
|
||||||
- name: Compile addon.node
|
|
||||||
run: npx cmake-js compile -T whisper-addon -B Release
|
|
||||||
|
|
||||||
- name: Download test model
|
|
||||||
run: |
|
|
||||||
bash ./models/download-ggml-model.sh base.en
|
|
||||||
- name: Test
|
|
||||||
run: |
|
|
||||||
cd examples/addon.node
|
|
||||||
npm run test
|
|
||||||
2
.gitignore
vendored
2
.gitignore
vendored
@ -1,5 +1,4 @@
|
|||||||
*.o
|
*.o
|
||||||
*.a
|
|
||||||
.cache/
|
.cache/
|
||||||
.vs/
|
.vs/
|
||||||
.vscode/
|
.vscode/
|
||||||
@ -19,7 +18,6 @@ build-sanitize-thread/
|
|||||||
/talk
|
/talk
|
||||||
/bench
|
/bench
|
||||||
|
|
||||||
arm_neon.h
|
|
||||||
sync.sh
|
sync.sh
|
||||||
libwhisper.a
|
libwhisper.a
|
||||||
libwhisper.so
|
libwhisper.so
|
||||||
|
|||||||
@ -1,6 +1,6 @@
|
|||||||
cmake_minimum_required (VERSION 3.0)
|
cmake_minimum_required (VERSION 3.0)
|
||||||
|
|
||||||
project(whisper.cpp VERSION 1.2.0)
|
project(whisper.cpp VERSION 1.1.0)
|
||||||
|
|
||||||
# Add path to modules
|
# Add path to modules
|
||||||
list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake/")
|
list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake/")
|
||||||
@ -226,13 +226,10 @@ target_compile_definitions(${TARGET} PUBLIC
|
|||||||
${WHISPER_EXTRA_FLAGS}
|
${WHISPER_EXTRA_FLAGS}
|
||||||
)
|
)
|
||||||
|
|
||||||
set_target_properties(${TARGET} PROPERTIES PUBLIC_HEADER "whisper.h")
|
|
||||||
|
|
||||||
install(TARGETS ${TARGET}
|
install(TARGETS ${TARGET}
|
||||||
LIBRARY DESTINATION lib
|
LIBRARY DESTINATION lib
|
||||||
ARCHIVE DESTINATION lib/static
|
ARCHIVE DESTINATION lib/static
|
||||||
RUNTIME DESTINATION bin
|
RUNTIME DESTINATION bin
|
||||||
PUBLIC_HEADER DESTINATION include
|
|
||||||
)
|
)
|
||||||
|
|
||||||
#
|
#
|
||||||
@ -245,7 +242,7 @@ add_subdirectory(bindings)
|
|||||||
# programs, examples and tests
|
# programs, examples and tests
|
||||||
#
|
#
|
||||||
|
|
||||||
if (WHISPER_BUILD_TESTS AND NOT CMAKE_JS_VERSION)
|
if (WHISPER_BUILD_TESTS)
|
||||||
enable_testing()
|
enable_testing()
|
||||||
add_subdirectory(tests)
|
add_subdirectory(tests)
|
||||||
endif ()
|
endif ()
|
||||||
|
|||||||
6
Makefile
6
Makefile
@ -115,15 +115,11 @@ endif
|
|||||||
ifeq ($(UNAME_M),amd64)
|
ifeq ($(UNAME_M),amd64)
|
||||||
CFLAGS += -mavx -mavx2 -mfma -mf16c
|
CFLAGS += -mavx -mavx2 -mfma -mf16c
|
||||||
endif
|
endif
|
||||||
ifneq ($(filter ppc64%,$(UNAME_M)),)
|
ifeq ($(UNAME_M),ppc64le)
|
||||||
POWER9_M := $(shell grep "POWER9" /proc/cpuinfo)
|
POWER9_M := $(shell grep "POWER9" /proc/cpuinfo)
|
||||||
ifneq (,$(findstring POWER9,$(POWER9_M)))
|
ifneq (,$(findstring POWER9,$(POWER9_M)))
|
||||||
CFLAGS += -mpower9-vector
|
CFLAGS += -mpower9-vector
|
||||||
endif
|
endif
|
||||||
# Require c++23's std::byteswap for big-endian support.
|
|
||||||
ifeq ($(UNAME_M),ppc64)
|
|
||||||
CXXFLAGS += -std=c++23 -DGGML_BIG_ENDIAN
|
|
||||||
endif
|
|
||||||
endif
|
endif
|
||||||
ifndef WHISPER_NO_ACCELERATE
|
ifndef WHISPER_NO_ACCELERATE
|
||||||
# Mac M1 - include Accelerate framework
|
# Mac M1 - include Accelerate framework
|
||||||
|
|||||||
138
README.md
138
README.md
@ -4,7 +4,7 @@
|
|||||||
[](https://opensource.org/licenses/MIT)
|
[](https://opensource.org/licenses/MIT)
|
||||||
[](https://www.npmjs.com/package/whisper.cpp/)
|
[](https://www.npmjs.com/package/whisper.cpp/)
|
||||||
|
|
||||||
Stable: [v1.2.0](https://github.com/ggerganov/whisper.cpp/releases/tag/v1.2.0) / [Roadmap | F.A.Q.](https://github.com/ggerganov/whisper.cpp/discussions/126)
|
Stable: [v1.0.4](https://github.com/ggerganov/whisper.cpp/releases/tag/v1.0.4) / Beta: [v1.1.0](https://github.com/ggerganov/whisper.cpp/releases/tag/v1.1.0) / [Roadmap | F.A.Q.](https://github.com/ggerganov/whisper.cpp/discussions/126)
|
||||||
|
|
||||||
High-performance inference of [OpenAI's Whisper](https://github.com/openai/whisper) automatic speech recognition (ASR) model:
|
High-performance inference of [OpenAI's Whisper](https://github.com/openai/whisper) automatic speech recognition (ASR) model:
|
||||||
|
|
||||||
@ -13,7 +13,7 @@ High-performance inference of [OpenAI's Whisper](https://github.com/openai/whisp
|
|||||||
- AVX intrinsics support for x86 architectures
|
- AVX intrinsics support for x86 architectures
|
||||||
- VSX intrinsics support for POWER architectures
|
- VSX intrinsics support for POWER architectures
|
||||||
- Mixed F16 / F32 precision
|
- Mixed F16 / F32 precision
|
||||||
- Low memory usage (Flash Attention)
|
- Low memory usage (Flash Attention + Flash Forward)
|
||||||
- Zero memory allocations at runtime
|
- Zero memory allocations at runtime
|
||||||
- Runs on the CPU
|
- Runs on the CPU
|
||||||
- [C-style API](https://github.com/ggerganov/whisper.cpp/blob/master/whisper.h)
|
- [C-style API](https://github.com/ggerganov/whisper.cpp/blob/master/whisper.h)
|
||||||
@ -105,13 +105,11 @@ options:
|
|||||||
-su, --speed-up [false ] speed up audio by x2 (reduced accuracy)
|
-su, --speed-up [false ] speed up audio by x2 (reduced accuracy)
|
||||||
-tr, --translate [false ] translate from source language to english
|
-tr, --translate [false ] translate from source language to english
|
||||||
-di, --diarize [false ] stereo audio diarization
|
-di, --diarize [false ] stereo audio diarization
|
||||||
-nf, --no-fallback [false ] do not use temperature fallback while decoding
|
|
||||||
-otxt, --output-txt [false ] output result in a text file
|
-otxt, --output-txt [false ] output result in a text file
|
||||||
-ovtt, --output-vtt [false ] output result in a vtt file
|
-ovtt, --output-vtt [false ] output result in a vtt file
|
||||||
-osrt, --output-srt [false ] output result in a srt file
|
-osrt, --output-srt [false ] output result in a srt file
|
||||||
-owts, --output-words [false ] output script for generating karaoke video
|
-owts, --output-words [false ] output script for generating karaoke video
|
||||||
-ocsv, --output-csv [false ] output result in a CSV file
|
-ocsv, --output-csv [false ] output result in a CSV file
|
||||||
-of FNAME, --output-file FNAME [ ] output file path (without file extension)
|
|
||||||
-ps, --print-special [false ] print special tokens
|
-ps, --print-special [false ] print special tokens
|
||||||
-pc, --print-colors [false ] print colors
|
-pc, --print-colors [false ] print colors
|
||||||
-pp, --print-progress [false ] print progress
|
-pp, --print-progress [false ] print progress
|
||||||
@ -139,8 +137,7 @@ Running base.en on all samples in ./samples ...
|
|||||||
[+] Running base.en on samples/jfk.wav ... (run 'ffplay samples/jfk.wav' to listen)
|
[+] Running base.en on samples/jfk.wav ... (run 'ffplay samples/jfk.wav' to listen)
|
||||||
----------------------------------------------
|
----------------------------------------------
|
||||||
|
|
||||||
whisper_init_from_file: loading model from 'models/ggml-base.en.bin'
|
whisper_model_load: loading model from 'models/ggml-base.en.bin'
|
||||||
whisper_model_load: loading model
|
|
||||||
whisper_model_load: n_vocab = 51864
|
whisper_model_load: n_vocab = 51864
|
||||||
whisper_model_load: n_audio_ctx = 1500
|
whisper_model_load: n_audio_ctx = 1500
|
||||||
whisper_model_load: n_audio_state = 512
|
whisper_model_load: n_audio_state = 512
|
||||||
@ -153,14 +150,13 @@ whisper_model_load: n_text_layer = 6
|
|||||||
whisper_model_load: n_mels = 80
|
whisper_model_load: n_mels = 80
|
||||||
whisper_model_load: f16 = 1
|
whisper_model_load: f16 = 1
|
||||||
whisper_model_load: type = 2
|
whisper_model_load: type = 2
|
||||||
whisper_model_load: mem required = 215.00 MB (+ 6.00 MB per decoder)
|
|
||||||
whisper_model_load: kv self size = 5.25 MB
|
|
||||||
whisper_model_load: kv cross size = 17.58 MB
|
|
||||||
whisper_model_load: adding 1607 extra tokens
|
whisper_model_load: adding 1607 extra tokens
|
||||||
whisper_model_load: model ctx = 140.60 MB
|
whisper_model_load: mem_required = 506.00 MB
|
||||||
|
whisper_model_load: ggml ctx size = 140.60 MB
|
||||||
|
whisper_model_load: memory size = 22.83 MB
|
||||||
whisper_model_load: model size = 140.54 MB
|
whisper_model_load: model size = 140.54 MB
|
||||||
|
|
||||||
system_info: n_threads = 4 / 10 | AVX = 0 | AVX2 = 0 | AVX512 = 0 | FMA = 0 | NEON = 1 | ARM_FMA = 1 | F16C = 0 | FP16_VA = 1 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 0 | VSX = 0 |
|
system_info: n_threads = 4 / 10 | AVX = 0 | AVX2 = 0 | AVX512 = 0 | NEON = 1 | FP16_VA = 1 | WASM_SIMD = 0 | BLAS = 1 |
|
||||||
|
|
||||||
main: processing 'samples/jfk.wav' (176000 samples, 11.0 sec), 4 threads, 1 processors, lang = en, task = transcribe, timestamps = 1 ...
|
main: processing 'samples/jfk.wav' (176000 samples, 11.0 sec), 4 threads, 1 processors, lang = en, task = transcribe, timestamps = 1 ...
|
||||||
|
|
||||||
@ -168,13 +164,12 @@ main: processing 'samples/jfk.wav' (176000 samples, 11.0 sec), 4 threads, 1 proc
|
|||||||
[00:00:00.000 --> 00:00:11.000] And so my fellow Americans, ask not what your country can do for you, ask what you can do for your country.
|
[00:00:00.000 --> 00:00:11.000] And so my fellow Americans, ask not what your country can do for you, ask what you can do for your country.
|
||||||
|
|
||||||
|
|
||||||
whisper_print_timings: fallbacks = 0 p / 0 h
|
whisper_print_timings: load time = 105.91 ms
|
||||||
whisper_print_timings: load time = 113.81 ms
|
whisper_print_timings: mel time = 24.62 ms
|
||||||
whisper_print_timings: mel time = 15.40 ms
|
whisper_print_timings: sample time = 3.63 ms
|
||||||
whisper_print_timings: sample time = 11.58 ms / 27 runs ( 0.43 ms per run)
|
whisper_print_timings: encode time = 324.71 ms / 54.12 ms per layer
|
||||||
whisper_print_timings: encode time = 266.60 ms / 1 runs ( 266.60 ms per run)
|
whisper_print_timings: decode time = 83.58 ms / 13.93 ms per layer
|
||||||
whisper_print_timings: decode time = 66.11 ms / 27 runs ( 2.45 ms per run)
|
whisper_print_timings: total time = 542.81 ms
|
||||||
whisper_print_timings: total time = 476.31 ms
|
|
||||||
```
|
```
|
||||||
|
|
||||||
The command downloads the `base.en` model converted to custom `ggml` format and runs the inference on all `.wav` samples in the folder `samples`.
|
The command downloads the `base.en` model converted to custom `ggml` format and runs the inference on all `.wav` samples in the folder `samples`.
|
||||||
@ -217,11 +212,11 @@ make large
|
|||||||
|
|
||||||
| Model | Disk | Mem | SHA |
|
| Model | Disk | Mem | SHA |
|
||||||
| --- | --- | --- | --- |
|
| --- | --- | --- | --- |
|
||||||
| tiny | 75 MB | ~125 MB | `bd577a113a864445d4c299885e0cb97d4ba92b5f` |
|
| tiny | 75 MB | ~390 MB | `bd577a113a864445d4c299885e0cb97d4ba92b5f` |
|
||||||
| base | 142 MB | ~210 MB | `465707469ff3a37a2b9b8d8f89f2f99de7299dac` |
|
| base | 142 MB | ~500 MB | `465707469ff3a37a2b9b8d8f89f2f99de7299dac` |
|
||||||
| small | 466 MB | ~600 MB | `55356645c2b361a969dfd0ef2c5a50d530afd8d5` |
|
| small | 466 MB | ~1.0 GB | `55356645c2b361a969dfd0ef2c5a50d530afd8d5` |
|
||||||
| medium | 1.5 GB | ~1.7 GB | `fd9727b6e1217c2f614f9b698455c4ffd82463b4` |
|
| medium | 1.5 GB | ~2.6 GB | `fd9727b6e1217c2f614f9b698455c4ffd82463b4` |
|
||||||
| large | 2.9 GB | ~3.3 GB | `0f4c8e34f21cf1a914c59d8b3ce882345ad349d6` |
|
| large | 2.9 GB | ~4.7 GB | `0f4c8e34f21cf1a914c59d8b3ce882345ad349d6` |
|
||||||
|
|
||||||
## Limitations
|
## Limitations
|
||||||
|
|
||||||
@ -239,8 +234,7 @@ in about half a minute on a MacBook M1 Pro, using `medium.en` model:
|
|||||||
```java
|
```java
|
||||||
$ ./main -m models/ggml-medium.en.bin -f samples/gb1.wav -t 8
|
$ ./main -m models/ggml-medium.en.bin -f samples/gb1.wav -t 8
|
||||||
|
|
||||||
whisper_init_from_file: loading model from 'models/ggml-medium.en.bin'
|
whisper_model_load: loading model from 'models/ggml-medium.en.bin'
|
||||||
whisper_model_load: loading model
|
|
||||||
whisper_model_load: n_vocab = 51864
|
whisper_model_load: n_vocab = 51864
|
||||||
whisper_model_load: n_audio_ctx = 1500
|
whisper_model_load: n_audio_ctx = 1500
|
||||||
whisper_model_load: n_audio_state = 1024
|
whisper_model_load: n_audio_state = 1024
|
||||||
@ -253,60 +247,55 @@ whisper_model_load: n_text_layer = 24
|
|||||||
whisper_model_load: n_mels = 80
|
whisper_model_load: n_mels = 80
|
||||||
whisper_model_load: f16 = 1
|
whisper_model_load: f16 = 1
|
||||||
whisper_model_load: type = 4
|
whisper_model_load: type = 4
|
||||||
whisper_model_load: mem required = 1720.00 MB (+ 43.00 MB per decoder)
|
whisper_model_load: mem_required = 2610.00 MB
|
||||||
whisper_model_load: kv self size = 42.00 MB
|
|
||||||
whisper_model_load: kv cross size = 140.62 MB
|
|
||||||
whisper_model_load: adding 1607 extra tokens
|
whisper_model_load: adding 1607 extra tokens
|
||||||
whisper_model_load: model ctx = 1462.35 MB
|
whisper_model_load: ggml ctx size = 1644.97 MB
|
||||||
|
whisper_model_load: memory size = 182.62 MB
|
||||||
whisper_model_load: model size = 1462.12 MB
|
whisper_model_load: model size = 1462.12 MB
|
||||||
|
|
||||||
system_info: n_threads = 8 / 10 | AVX = 0 | AVX2 = 0 | AVX512 = 0 | FMA = 0 | NEON = 1 | ARM_FMA = 1 | F16C = 0 | FP16_VA = 1 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 0 | VSX = 0 |
|
main: processing 'samples/gb1.wav' (3179750 samples, 198.7 sec), 8 threads, lang = en, task = transcribe, timestamps = 1 ...
|
||||||
|
|
||||||
main: processing 'samples/gb1.wav' (3179750 samples, 198.7 sec), 8 threads, 1 processors, lang = en, task = transcribe, timestamps = 1 ...
|
[00:00.000 --> 00:08.000] My fellow Americans, this day has brought terrible news and great sadness to our country.
|
||||||
|
[00:08.000 --> 00:17.000] At nine o'clock this morning, Mission Control in Houston lost contact with our Space Shuttle Columbia.
|
||||||
|
[00:17.000 --> 00:23.000] A short time later, debris was seen falling from the skies above Texas.
|
||||||
|
[00:23.000 --> 00:29.000] The Columbia's lost. There are no survivors.
|
||||||
|
[00:29.000 --> 00:32.000] On board was a crew of seven.
|
||||||
|
[00:32.000 --> 00:39.000] Colonel Rick Husband, Lieutenant Colonel Michael Anderson, Commander Laurel Clark,
|
||||||
|
[00:39.000 --> 00:48.000] Captain David Brown, Commander William McCool, Dr. Kultna Shavla, and Ilan Ramon,
|
||||||
|
[00:48.000 --> 00:52.000] a colonel in the Israeli Air Force.
|
||||||
|
[00:52.000 --> 00:58.000] These men and women assumed great risk in the service to all humanity.
|
||||||
|
[00:58.000 --> 01:03.000] In an age when space flight has come to seem almost routine,
|
||||||
|
[01:03.000 --> 01:07.000] it is easy to overlook the dangers of travel by rocket
|
||||||
|
[01:07.000 --> 01:12.000] and the difficulties of navigating the fierce outer atmosphere of the Earth.
|
||||||
|
[01:12.000 --> 01:18.000] These astronauts knew the dangers, and they faced them willingly,
|
||||||
|
[01:18.000 --> 01:23.000] knowing they had a high and noble purpose in life.
|
||||||
|
[01:23.000 --> 01:31.000] Because of their courage and daring and idealism, we will miss them all the more.
|
||||||
|
[01:31.000 --> 01:36.000] All Americans today are thinking as well of the families of these men and women
|
||||||
|
[01:36.000 --> 01:40.000] who have been given this sudden shock and grief.
|
||||||
|
[01:40.000 --> 01:45.000] You're not alone. Our entire nation grieves with you,
|
||||||
|
[01:45.000 --> 01:52.000] and those you love will always have the respect and gratitude of this country.
|
||||||
|
[01:52.000 --> 01:56.000] The cause in which they died will continue.
|
||||||
|
[01:56.000 --> 02:04.000] Mankind is led into the darkness beyond our world by the inspiration of discovery
|
||||||
|
[02:04.000 --> 02:11.000] and the longing to understand. Our journey into space will go on.
|
||||||
|
[02:11.000 --> 02:16.000] In the skies today, we saw destruction and tragedy.
|
||||||
|
[02:16.000 --> 02:22.000] Yet farther than we can see, there is comfort and hope.
|
||||||
|
[02:22.000 --> 02:29.000] In the words of the prophet Isaiah, "Lift your eyes and look to the heavens
|
||||||
|
[02:29.000 --> 02:35.000] who created all these. He who brings out the starry hosts one by one
|
||||||
|
[02:35.000 --> 02:39.000] and calls them each by name."
|
||||||
|
[02:39.000 --> 02:46.000] Because of His great power and mighty strength, not one of them is missing.
|
||||||
|
[02:46.000 --> 02:55.000] The same Creator who names the stars also knows the names of the seven souls we mourn today.
|
||||||
|
[02:55.000 --> 03:01.000] The crew of the shuttle Columbia did not return safely to earth,
|
||||||
|
[03:01.000 --> 03:05.000] yet we can pray that all are safely home.
|
||||||
|
[03:05.000 --> 03:13.000] May God bless the grieving families, and may God continue to bless America.
|
||||||
|
[03:13.000 --> 03:41.000] Audio
|
||||||
|
|
||||||
|
|
||||||
[00:00:00.000 --> 00:00:08.000] My fellow Americans, this day has brought terrible news and great sadness to our country.
|
whisper_print_timings: load time = 575.92 ms
|
||||||
[00:00:08.000 --> 00:00:17.000] At nine o'clock this morning, Mission Control in Houston lost contact with our Space Shuttle Columbia.
|
whisper_print_timings: mel time = 230.60 ms
|
||||||
[00:00:17.000 --> 00:00:23.000] A short time later, debris was seen falling from the skies above Texas.
|
whisper_print_timings: sample time = 73.19 ms
|
||||||
[00:00:23.000 --> 00:00:29.000] The Columbia's lost. There are no survivors.
|
whisper_print_timings: encode time = 19552.61 ms / 814.69 ms per layer
|
||||||
[00:00:29.000 --> 00:00:32.000] On board was a crew of seven.
|
whisper_print_timings: decode time = 13249.96 ms / 552.08 ms per layer
|
||||||
[00:00:32.000 --> 00:00:39.000] Colonel Rick Husband, Lieutenant Colonel Michael Anderson, Commander Laurel Clark,
|
whisper_print_timings: total time = 33686.27 ms
|
||||||
[00:00:39.000 --> 00:00:48.000] Captain David Brown, Commander William McCool, Dr. Kultna Shavla, and Ilan Ramon,
|
|
||||||
[00:00:48.000 --> 00:00:52.000] a colonel in the Israeli Air Force.
|
|
||||||
[00:00:52.000 --> 00:00:58.000] These men and women assumed great risk in the service to all humanity.
|
|
||||||
[00:00:58.000 --> 00:01:03.000] In an age when space flight has come to seem almost routine,
|
|
||||||
[00:01:03.000 --> 00:01:07.000] it is easy to overlook the dangers of travel by rocket
|
|
||||||
[00:01:07.000 --> 00:01:12.000] and the difficulties of navigating the fierce outer atmosphere of the Earth.
|
|
||||||
[00:01:12.000 --> 00:01:18.000] These astronauts knew the dangers, and they faced them willingly,
|
|
||||||
[00:01:18.000 --> 00:01:23.000] knowing they had a high and noble purpose in life.
|
|
||||||
[00:01:23.000 --> 00:01:31.000] Because of their courage and daring and idealism, we will miss them all the more.
|
|
||||||
[00:01:31.000 --> 00:01:36.000] All Americans today are thinking as well of the families of these men and women
|
|
||||||
[00:01:36.000 --> 00:01:40.000] who have been given this sudden shock and grief.
|
|
||||||
[00:01:40.000 --> 00:01:45.000] You're not alone. Our entire nation grieves with you,
|
|
||||||
[00:01:45.000 --> 00:01:52.000] and those you love will always have the respect and gratitude of this country.
|
|
||||||
[00:01:52.000 --> 00:01:56.000] The cause in which they died will continue.
|
|
||||||
[00:01:56.000 --> 00:02:04.000] Mankind is led into the darkness beyond our world by the inspiration of discovery
|
|
||||||
[00:02:04.000 --> 00:02:11.000] and the longing to understand. Our journey into space will go on.
|
|
||||||
[00:02:11.000 --> 00:02:16.000] In the skies today, we saw destruction and tragedy.
|
|
||||||
[00:02:16.000 --> 00:02:22.000] Yet farther than we can see, there is comfort and hope.
|
|
||||||
[00:02:22.000 --> 00:02:29.000] In the words of the prophet Isaiah, "Lift your eyes and look to the heavens
|
|
||||||
[00:02:29.000 --> 00:02:35.000] who created all these. He who brings out the starry hosts one by one
|
|
||||||
[00:02:35.000 --> 00:02:39.000] and calls them each by name."
|
|
||||||
[00:02:39.000 --> 00:02:46.000] Because of His great power and mighty strength, not one of them is missing.
|
|
||||||
[00:02:46.000 --> 00:02:55.000] The same Creator who names the stars also knows the names of the seven souls we mourn today.
|
|
||||||
[00:02:55.000 --> 00:03:01.000] The crew of the shuttle Columbia did not return safely to earth,
|
|
||||||
[00:03:01.000 --> 00:03:05.000] yet we can pray that all are safely home.
|
|
||||||
[00:03:05.000 --> 00:03:13.000] May God bless the grieving families, and may God continue to bless America.
|
|
||||||
[00:03:13.000 --> 00:03:19.000] [Silence]
|
|
||||||
|
|
||||||
|
|
||||||
whisper_print_timings: fallbacks = 1 p / 0 h
|
|
||||||
whisper_print_timings: load time = 569.03 ms
|
|
||||||
whisper_print_timings: mel time = 146.85 ms
|
|
||||||
whisper_print_timings: sample time = 238.66 ms / 553 runs ( 0.43 ms per run)
|
|
||||||
whisper_print_timings: encode time = 18665.10 ms / 9 runs ( 2073.90 ms per run)
|
|
||||||
whisper_print_timings: decode time = 13090.93 ms / 549 runs ( 23.85 ms per run)
|
|
||||||
whisper_print_timings: total time = 32733.52 ms
|
|
||||||
```
|
```
|
||||||
</details>
|
</details>
|
||||||
|
|
||||||
@ -465,9 +454,6 @@ in [models](models).
|
|||||||
- [X] Javascript: [bindings/javascript](bindings/javascript) | [#309](https://github.com/ggerganov/whisper.cpp/discussions/309)
|
- [X] Javascript: [bindings/javascript](bindings/javascript) | [#309](https://github.com/ggerganov/whisper.cpp/discussions/309)
|
||||||
- [X] Go: [bindings/go](bindings/go) | [#312](https://github.com/ggerganov/whisper.cpp/discussions/312)
|
- [X] Go: [bindings/go](bindings/go) | [#312](https://github.com/ggerganov/whisper.cpp/discussions/312)
|
||||||
- [X] Objective-C / Swift: [ggerganov/whisper.spm](https://github.com/ggerganov/whisper.spm) | [#313](https://github.com/ggerganov/whisper.cpp/discussions/313)
|
- [X] Objective-C / Swift: [ggerganov/whisper.spm](https://github.com/ggerganov/whisper.spm) | [#313](https://github.com/ggerganov/whisper.cpp/discussions/313)
|
||||||
- [X] .NET:
|
|
||||||
- [sandrohanea/whisper.net](https://github.com/sandrohanea/whisper.net)
|
|
||||||
- [NickDarvey/whisper](https://github.com/NickDarvey/whisper)
|
|
||||||
- [ ] Python: soon | [WIP](https://github.com/ggerganov/whisper.cpp/issues/9)
|
- [ ] Python: soon | [WIP](https://github.com/ggerganov/whisper.cpp/issues/9)
|
||||||
|
|
||||||
## Examples
|
## Examples
|
||||||
|
|||||||
@ -25,8 +25,6 @@ func Process(model whisper.Model, path string, flags *Flags) error {
|
|||||||
return err
|
return err
|
||||||
}
|
}
|
||||||
|
|
||||||
fmt.Printf("\n%s\n", context.SystemInfo())
|
|
||||||
|
|
||||||
// Open the file
|
// Open the file
|
||||||
fmt.Fprintf(flags.Output(), "Loading %q\n", path)
|
fmt.Fprintf(flags.Output(), "Loading %q\n", path)
|
||||||
fh, err := os.Open(path)
|
fh, err := os.Open(path)
|
||||||
@ -66,13 +64,10 @@ func Process(model whisper.Model, path string, flags *Flags) error {
|
|||||||
|
|
||||||
// Process the data
|
// Process the data
|
||||||
fmt.Fprintf(flags.Output(), " ...processing %q\n", path)
|
fmt.Fprintf(flags.Output(), " ...processing %q\n", path)
|
||||||
context.ResetTimings()
|
|
||||||
if err := context.Process(data, cb); err != nil {
|
if err := context.Process(data, cb); err != nil {
|
||||||
return err
|
return err
|
||||||
}
|
}
|
||||||
|
|
||||||
context.PrintTimings()
|
|
||||||
|
|
||||||
// Print out the results
|
// Print out the results
|
||||||
switch {
|
switch {
|
||||||
case flags.GetOut() == "srt":
|
case flags.GetOut() == "srt":
|
||||||
|
|||||||
@ -49,10 +49,6 @@ func (p *Params) SetSpeedup(v bool) {
|
|||||||
|
|
||||||
// Set language id
|
// Set language id
|
||||||
func (p *Params) SetLanguage(lang int) error {
|
func (p *Params) SetLanguage(lang int) error {
|
||||||
if lang == -1 {
|
|
||||||
p.language = nil
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
str := C.whisper_lang_str(C.int(lang))
|
str := C.whisper_lang_str(C.int(lang))
|
||||||
if str == nil {
|
if str == nil {
|
||||||
return ErrInvalidLanguage
|
return ErrInvalidLanguage
|
||||||
@ -70,11 +66,6 @@ func (p *Params) Language() int {
|
|||||||
return int(C.whisper_lang_id(p.language))
|
return int(C.whisper_lang_id(p.language))
|
||||||
}
|
}
|
||||||
|
|
||||||
// Threads available
|
|
||||||
func (p *Params) Threads() int {
|
|
||||||
return int(p.n_threads)
|
|
||||||
}
|
|
||||||
|
|
||||||
// Set number of threads to use
|
// Set number of threads to use
|
||||||
func (p *Params) SetThreads(threads int) {
|
func (p *Params) SetThreads(threads int) {
|
||||||
p.n_threads = C.int(threads)
|
p.n_threads = C.int(threads)
|
||||||
|
|||||||
@ -1,9 +1,7 @@
|
|||||||
package whisper
|
package whisper
|
||||||
|
|
||||||
import (
|
import (
|
||||||
"fmt"
|
|
||||||
"io"
|
"io"
|
||||||
"runtime"
|
|
||||||
"strings"
|
"strings"
|
||||||
"time"
|
"time"
|
||||||
|
|
||||||
@ -46,10 +44,7 @@ func (context *context) SetLanguage(lang string) error {
|
|||||||
if !context.model.IsMultilingual() {
|
if !context.model.IsMultilingual() {
|
||||||
return ErrModelNotMultilingual
|
return ErrModelNotMultilingual
|
||||||
}
|
}
|
||||||
|
if id := context.model.ctx.Whisper_lang_id(lang); id < 0 {
|
||||||
if lang == "auto" {
|
|
||||||
context.params.SetLanguage(-1)
|
|
||||||
} else if id := context.model.ctx.Whisper_lang_id(lang); id < 0 {
|
|
||||||
return ErrUnsupportedLanguage
|
return ErrUnsupportedLanguage
|
||||||
} else if err := context.params.SetLanguage(id); err != nil {
|
} else if err := context.params.SetLanguage(id); err != nil {
|
||||||
return err
|
return err
|
||||||
@ -64,10 +59,6 @@ func (context *context) IsMultilingual() bool {
|
|||||||
|
|
||||||
// Get language
|
// Get language
|
||||||
func (context *context) Language() string {
|
func (context *context) Language() string {
|
||||||
id := context.params.Language()
|
|
||||||
if id == -1 {
|
|
||||||
return "auto"
|
|
||||||
}
|
|
||||||
return whisper.Whisper_lang_str(context.params.Language())
|
return whisper.Whisper_lang_str(context.params.Language())
|
||||||
}
|
}
|
||||||
|
|
||||||
@ -116,36 +107,6 @@ func (context *context) SetMaxTokensPerSegment(n uint) {
|
|||||||
context.params.SetMaxTokensPerSegment(int(n))
|
context.params.SetMaxTokensPerSegment(int(n))
|
||||||
}
|
}
|
||||||
|
|
||||||
// ResetTimings resets the mode timings. Should be called before processing
|
|
||||||
func (context *context) ResetTimings() {
|
|
||||||
context.model.ctx.Whisper_reset_timings()
|
|
||||||
}
|
|
||||||
|
|
||||||
// PrintTimings prints the model timings to stdout.
|
|
||||||
func (context *context) PrintTimings() {
|
|
||||||
context.model.ctx.Whisper_print_timings()
|
|
||||||
}
|
|
||||||
|
|
||||||
// SystemInfo returns the system information
|
|
||||||
func (context *context) SystemInfo() string {
|
|
||||||
return fmt.Sprintf("system_info: n_threads = %d / %d | %s\n",
|
|
||||||
context.params.Threads(),
|
|
||||||
runtime.NumCPU(),
|
|
||||||
whisper.Whisper_print_system_info(),
|
|
||||||
)
|
|
||||||
}
|
|
||||||
|
|
||||||
// Use mel data at offset_ms to try and auto-detect the spoken language
|
|
||||||
// Make sure to call whisper_pcm_to_mel() or whisper_set_mel() first.
|
|
||||||
// Returns the probabilities of all languages.
|
|
||||||
func (context *context) WhisperLangAutoDetect(offset_ms int, n_threads int) ([]float32, error) {
|
|
||||||
langProbs, err := context.model.ctx.Whisper_lang_auto_detect(offset_ms, n_threads)
|
|
||||||
if err != nil {
|
|
||||||
return nil, err
|
|
||||||
}
|
|
||||||
return langProbs, nil
|
|
||||||
}
|
|
||||||
|
|
||||||
// Process new sample data and return any errors
|
// Process new sample data and return any errors
|
||||||
func (context *context) Process(data []float32, cb SegmentCallback) error {
|
func (context *context) Process(data []float32, cb SegmentCallback) error {
|
||||||
if context.model.ctx == nil {
|
if context.model.ctx == nil {
|
||||||
|
|||||||
@ -29,7 +29,7 @@ type Model interface {
|
|||||||
|
|
||||||
// Context is the speach recognition context.
|
// Context is the speach recognition context.
|
||||||
type Context interface {
|
type Context interface {
|
||||||
SetLanguage(string) error // Set the language to use for speech recognition, use "auto" for auto detect language.
|
SetLanguage(string) error // Set the language to use for speech recognition.
|
||||||
SetTranslate(bool) // Set translate flag
|
SetTranslate(bool) // Set translate flag
|
||||||
IsMultilingual() bool // Return true if the model is multilingual.
|
IsMultilingual() bool // Return true if the model is multilingual.
|
||||||
Language() string // Get language
|
Language() string // Get language
|
||||||
@ -60,12 +60,6 @@ type Context interface {
|
|||||||
IsNOT(Token) bool // Test for "No timestamps" token
|
IsNOT(Token) bool // Test for "No timestamps" token
|
||||||
IsLANG(Token, string) bool // Test for token associated with a specific language
|
IsLANG(Token, string) bool // Test for token associated with a specific language
|
||||||
IsText(Token) bool // Test for text token
|
IsText(Token) bool // Test for text token
|
||||||
|
|
||||||
// Timings
|
|
||||||
PrintTimings()
|
|
||||||
ResetTimings()
|
|
||||||
|
|
||||||
SystemInfo() string
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// Segment is the text result of a speech recognition.
|
// Segment is the text result of a speech recognition.
|
||||||
|
|||||||
Submodule bindings/ios updated: d5c6d5c8a3...f6334b026f
@ -1,6 +1,6 @@
|
|||||||
{
|
{
|
||||||
"name": "whisper.cpp",
|
"name": "whisper.cpp",
|
||||||
"version": "1.2.0",
|
"version": "1.1.0",
|
||||||
"description": "Whisper speech recognition",
|
"description": "Whisper speech recognition",
|
||||||
"main": "whisper.js",
|
"main": "whisper.js",
|
||||||
"scripts": {
|
"scripts": {
|
||||||
|
|||||||
File diff suppressed because one or more lines are too long
@ -24,8 +24,6 @@ if (EMSCRIPTEN)
|
|||||||
add_subdirectory(command.wasm)
|
add_subdirectory(command.wasm)
|
||||||
add_subdirectory(talk.wasm)
|
add_subdirectory(talk.wasm)
|
||||||
add_subdirectory(bench.wasm)
|
add_subdirectory(bench.wasm)
|
||||||
elseif(CMAKE_JS_VERSION)
|
|
||||||
add_subdirectory(addon.node)
|
|
||||||
else()
|
else()
|
||||||
add_subdirectory(main)
|
add_subdirectory(main)
|
||||||
add_subdirectory(stream)
|
add_subdirectory(stream)
|
||||||
|
|||||||
3
examples/addon.node/.gitignore
vendored
3
examples/addon.node/.gitignore
vendored
@ -1,3 +0,0 @@
|
|||||||
.idea
|
|
||||||
node_modules
|
|
||||||
build
|
|
||||||
@ -1,31 +0,0 @@
|
|||||||
set(TARGET whisper-addon)
|
|
||||||
|
|
||||||
# Base settings
|
|
||||||
#==================================================================
|
|
||||||
# env var supported by cmake-js
|
|
||||||
add_definitions(-DNAPI_VERSION=4)
|
|
||||||
include_directories(${CMAKE_JS_INC})
|
|
||||||
#==================================================================
|
|
||||||
|
|
||||||
add_library(${TARGET} SHARED ${CMAKE_JS_SRC} addon.cpp)
|
|
||||||
set_target_properties(${TARGET} PROPERTIES PREFIX "" SUFFIX ".node")
|
|
||||||
|
|
||||||
include(DefaultTargetOptions)
|
|
||||||
|
|
||||||
# Include N-API wrappers
|
|
||||||
#==================================================================
|
|
||||||
execute_process(COMMAND node -p "require('node-addon-api').include"
|
|
||||||
WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}
|
|
||||||
OUTPUT_VARIABLE NODE_ADDON_API_DIR
|
|
||||||
)
|
|
||||||
string(REPLACE "\n" "" NODE_ADDON_API_DIR ${NODE_ADDON_API_DIR})
|
|
||||||
string(REPLACE "\"" "" NODE_ADDON_API_DIR ${NODE_ADDON_API_DIR})
|
|
||||||
target_include_directories(${TARGET} PRIVATE ${NODE_ADDON_API_DIR})
|
|
||||||
#==================================================================
|
|
||||||
|
|
||||||
target_link_libraries(${TARGET} ${CMAKE_JS_LIB} whisper ${CMAKE_THREAD_LIBS_INIT})
|
|
||||||
|
|
||||||
if(MSVC AND CMAKE_JS_NODELIB_DEF AND CMAKE_JS_NODELIB_TARGET)
|
|
||||||
# Generate node.lib
|
|
||||||
execute_process(COMMAND ${CMAKE_AR} /def:${CMAKE_JS_NODELIB_DEF} /out:${CMAKE_JS_NODELIB_TARGET} ${CMAKE_STATIC_LINKER_FLAGS})
|
|
||||||
endif()
|
|
||||||
@ -1,37 +0,0 @@
|
|||||||
# addon
|
|
||||||
|
|
||||||
This is an addon demo that can **perform whisper model reasoning in `node` and `electron` environments**, based on [cmake-js](https://github.com/cmake-js/cmake-js).
|
|
||||||
It can be used as a reference for using the whisper.cpp project in other node projects.
|
|
||||||
|
|
||||||
## Install
|
|
||||||
|
|
||||||
```shell
|
|
||||||
npm install
|
|
||||||
```
|
|
||||||
|
|
||||||
## Compile
|
|
||||||
|
|
||||||
Make sure it is in the project root directory and compiled with make-js.
|
|
||||||
|
|
||||||
```shell
|
|
||||||
npx cmake-js compile -T whisper-addon -B Release
|
|
||||||
```
|
|
||||||
|
|
||||||
For Electron addon and cmake-js options, you can see [cmake-js](https://github.com/cmake-js/cmake-js) and make very few configuration changes.
|
|
||||||
|
|
||||||
> Such as appointing special cmake path:
|
|
||||||
> ```shell
|
|
||||||
> npx cmake-js compile -c 'xxx/cmake' -T whisper-addon -B Release
|
|
||||||
> ```
|
|
||||||
|
|
||||||
## Run
|
|
||||||
|
|
||||||
```shell
|
|
||||||
cd examples/addon.node
|
|
||||||
|
|
||||||
node index.js --language='language' --model='model-path' --fname_inp='file-path'
|
|
||||||
```
|
|
||||||
|
|
||||||
Because this is a simple Demo, only the above parameters are set in the node environment.
|
|
||||||
|
|
||||||
Other parameters can also be specified in the node environment.
|
|
||||||
@ -1,15 +0,0 @@
|
|||||||
const path = require('path');
|
|
||||||
const { whisper } = require(path.join(__dirname, '../../../build/Release/whisper-addon'));
|
|
||||||
|
|
||||||
const whisperParamsMock = {
|
|
||||||
language: 'en',
|
|
||||||
model: path.join(__dirname, '../../../models/ggml-base.en.bin'),
|
|
||||||
fname_inp: path.join(__dirname, '../../../samples/jfk.wav'),
|
|
||||||
};
|
|
||||||
|
|
||||||
describe("Run whisper.node", () => {
|
|
||||||
|
|
||||||
test("it should receive a non-empty value", () => {
|
|
||||||
expect(whisper(whisperParamsMock).length).toBeGreaterThan(0);
|
|
||||||
});
|
|
||||||
});
|
|
||||||
@ -1,422 +0,0 @@
|
|||||||
#include <cstdint>
|
|
||||||
#include <string>
|
|
||||||
#include <thread>
|
|
||||||
#include <vector>
|
|
||||||
#include <cmath>
|
|
||||||
|
|
||||||
#include "napi.h"
|
|
||||||
|
|
||||||
#define DR_WAV_IMPLEMENTATION
|
|
||||||
#include "dr_wav.h"
|
|
||||||
|
|
||||||
#include "whisper.h"
|
|
||||||
|
|
||||||
struct whisper_params {
|
|
||||||
int32_t n_threads = std::min(4, (int32_t) std::thread::hardware_concurrency());
|
|
||||||
int32_t n_processors = 1;
|
|
||||||
int32_t offset_t_ms = 0;
|
|
||||||
int32_t offset_n = 0;
|
|
||||||
int32_t duration_ms = 0;
|
|
||||||
int32_t max_context = -1;
|
|
||||||
int32_t max_len = 0;
|
|
||||||
int32_t best_of = 5;
|
|
||||||
int32_t beam_size = -1;
|
|
||||||
|
|
||||||
float word_thold = 0.01f;
|
|
||||||
float entropy_thold = 2.4f;
|
|
||||||
float logprob_thold = -1.0f;
|
|
||||||
|
|
||||||
bool speed_up = false;
|
|
||||||
bool translate = false;
|
|
||||||
bool diarize = false;
|
|
||||||
bool output_txt = false;
|
|
||||||
bool output_vtt = false;
|
|
||||||
bool output_srt = false;
|
|
||||||
bool output_wts = false;
|
|
||||||
bool output_csv = false;
|
|
||||||
bool print_special = false;
|
|
||||||
bool print_colors = false;
|
|
||||||
bool print_progress = false;
|
|
||||||
bool no_timestamps = false;
|
|
||||||
|
|
||||||
std::string language = "en";
|
|
||||||
std::string prompt;
|
|
||||||
std::string model = "../../ggml-large.bin";
|
|
||||||
|
|
||||||
std::vector<std::string> fname_inp = {};
|
|
||||||
std::vector<std::string> fname_outp = {};
|
|
||||||
};
|
|
||||||
|
|
||||||
struct whisper_print_user_data {
|
|
||||||
const whisper_params * params;
|
|
||||||
|
|
||||||
const std::vector<std::vector<float>> * pcmf32s;
|
|
||||||
};
|
|
||||||
|
|
||||||
// 500 -> 00:05.000
|
|
||||||
// 6000 -> 01:00.000
|
|
||||||
std::string to_timestamp(int64_t t, bool comma = false) {
|
|
||||||
int64_t msec = t * 10;
|
|
||||||
int64_t hr = msec / (1000 * 60 * 60);
|
|
||||||
msec = msec - hr * (1000 * 60 * 60);
|
|
||||||
int64_t min = msec / (1000 * 60);
|
|
||||||
msec = msec - min * (1000 * 60);
|
|
||||||
int64_t sec = msec / 1000;
|
|
||||||
msec = msec - sec * 1000;
|
|
||||||
|
|
||||||
char buf[32];
|
|
||||||
snprintf(buf, sizeof(buf), "%02d:%02d:%02d%s%03d", (int) hr, (int) min, (int) sec, comma ? "," : ".", (int) msec);
|
|
||||||
|
|
||||||
return std::string(buf);
|
|
||||||
}
|
|
||||||
|
|
||||||
int timestamp_to_sample(int64_t t, int n_samples) {
|
|
||||||
return std::max(0, std::min((int) n_samples - 1, (int) ((t*WHISPER_SAMPLE_RATE)/100)));
|
|
||||||
}
|
|
||||||
|
|
||||||
void whisper_print_segment_callback(struct whisper_context * ctx, int n_new, void * user_data) {
|
|
||||||
const auto & params = *((whisper_print_user_data *) user_data)->params;
|
|
||||||
const auto & pcmf32s = *((whisper_print_user_data *) user_data)->pcmf32s;
|
|
||||||
|
|
||||||
const int n_segments = whisper_full_n_segments(ctx);
|
|
||||||
|
|
||||||
std::string speaker = "";
|
|
||||||
|
|
||||||
int64_t t0;
|
|
||||||
int64_t t1;
|
|
||||||
|
|
||||||
// print the last n_new segments
|
|
||||||
const int s0 = n_segments - n_new;
|
|
||||||
|
|
||||||
if (s0 == 0) {
|
|
||||||
printf("\n");
|
|
||||||
}
|
|
||||||
|
|
||||||
for (int i = s0; i < n_segments; i++) {
|
|
||||||
if (!params.no_timestamps || params.diarize) {
|
|
||||||
t0 = whisper_full_get_segment_t0(ctx, i);
|
|
||||||
t1 = whisper_full_get_segment_t1(ctx, i);
|
|
||||||
}
|
|
||||||
|
|
||||||
if (!params.no_timestamps) {
|
|
||||||
printf("[%s --> %s] ", to_timestamp(t0).c_str(), to_timestamp(t1).c_str());
|
|
||||||
}
|
|
||||||
|
|
||||||
if (params.diarize && pcmf32s.size() == 2) {
|
|
||||||
const int64_t n_samples = pcmf32s[0].size();
|
|
||||||
|
|
||||||
const int64_t is0 = timestamp_to_sample(t0, n_samples);
|
|
||||||
const int64_t is1 = timestamp_to_sample(t1, n_samples);
|
|
||||||
|
|
||||||
double energy0 = 0.0f;
|
|
||||||
double energy1 = 0.0f;
|
|
||||||
|
|
||||||
for (int64_t j = is0; j < is1; j++) {
|
|
||||||
energy0 += fabs(pcmf32s[0][j]);
|
|
||||||
energy1 += fabs(pcmf32s[1][j]);
|
|
||||||
}
|
|
||||||
|
|
||||||
if (energy0 > 1.1*energy1) {
|
|
||||||
speaker = "(speaker 0)";
|
|
||||||
} else if (energy1 > 1.1*energy0) {
|
|
||||||
speaker = "(speaker 1)";
|
|
||||||
} else {
|
|
||||||
speaker = "(speaker ?)";
|
|
||||||
}
|
|
||||||
|
|
||||||
//printf("is0 = %lld, is1 = %lld, energy0 = %f, energy1 = %f, %s\n", is0, is1, energy0, energy1, speaker.c_str());
|
|
||||||
}
|
|
||||||
|
|
||||||
// colorful print bug
|
|
||||||
//
|
|
||||||
const char * text = whisper_full_get_segment_text(ctx, i);
|
|
||||||
printf("%s%s", speaker.c_str(), text);
|
|
||||||
|
|
||||||
|
|
||||||
// with timestamps or speakers: each segment on new line
|
|
||||||
if (!params.no_timestamps || params.diarize) {
|
|
||||||
printf("\n");
|
|
||||||
}
|
|
||||||
|
|
||||||
fflush(stdout);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
int run(whisper_params ¶ms, std::vector<std::vector<std::string>> &result) {
|
|
||||||
|
|
||||||
if (params.fname_inp.empty()) {
|
|
||||||
fprintf(stderr, "error: no input files specified\n");
|
|
||||||
return 2;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (params.language != "auto" && whisper_lang_id(params.language.c_str()) == -1) {
|
|
||||||
fprintf(stderr, "error: unknown language '%s'\n", params.language.c_str());
|
|
||||||
exit(0);
|
|
||||||
}
|
|
||||||
|
|
||||||
// whisper init
|
|
||||||
|
|
||||||
struct whisper_context * ctx = whisper_init_from_file(params.model.c_str());
|
|
||||||
|
|
||||||
if (ctx == nullptr) {
|
|
||||||
fprintf(stderr, "error: failed to initialize whisper context\n");
|
|
||||||
return 3;
|
|
||||||
}
|
|
||||||
|
|
||||||
// initial prompt
|
|
||||||
std::vector<whisper_token> prompt_tokens;
|
|
||||||
|
|
||||||
if (!params.prompt.empty()) {
|
|
||||||
prompt_tokens.resize(1024);
|
|
||||||
prompt_tokens.resize(whisper_tokenize(ctx, params.prompt.c_str(), prompt_tokens.data(), prompt_tokens.size()));
|
|
||||||
|
|
||||||
fprintf(stderr, "\n");
|
|
||||||
fprintf(stderr, "initial prompt: '%s'\n", params.prompt.c_str());
|
|
||||||
fprintf(stderr, "initial tokens: [ ");
|
|
||||||
for (int i = 0; i < (int) prompt_tokens.size(); ++i) {
|
|
||||||
fprintf(stderr, "%d ", prompt_tokens[i]);
|
|
||||||
}
|
|
||||||
fprintf(stderr, "]\n");
|
|
||||||
}
|
|
||||||
|
|
||||||
for (int f = 0; f < (int) params.fname_inp.size(); ++f) {
|
|
||||||
const auto fname_inp = params.fname_inp[f];
|
|
||||||
const auto fname_outp = f < (int)params.fname_outp.size() && !params.fname_outp[f].empty() ? params.fname_outp[f] : params.fname_inp[f];
|
|
||||||
|
|
||||||
std::vector<float> pcmf32; // mono-channel F32 PCM
|
|
||||||
std::vector<std::vector<float>> pcmf32s; // stereo-channel F32 PCM
|
|
||||||
|
|
||||||
// WAV input
|
|
||||||
{
|
|
||||||
drwav wav;
|
|
||||||
std::vector<uint8_t> wav_data; // used for pipe input from stdin
|
|
||||||
|
|
||||||
if (fname_inp == "-") {
|
|
||||||
{
|
|
||||||
uint8_t buf[1024];
|
|
||||||
while (true)
|
|
||||||
{
|
|
||||||
const size_t n = fread(buf, 1, sizeof(buf), stdin);
|
|
||||||
if (n == 0) {
|
|
||||||
break;
|
|
||||||
}
|
|
||||||
wav_data.insert(wav_data.end(), buf, buf + n);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
if (drwav_init_memory(&wav, wav_data.data(), wav_data.size(), nullptr) == false) {
|
|
||||||
fprintf(stderr, "error: failed to open WAV file from stdin\n");
|
|
||||||
return 4;
|
|
||||||
}
|
|
||||||
|
|
||||||
fprintf(stderr, "%s: read %zu bytes from stdin\n", __func__, wav_data.size());
|
|
||||||
}
|
|
||||||
else if (drwav_init_file(&wav, fname_inp.c_str(), nullptr) == false) {
|
|
||||||
fprintf(stderr, "error: failed to open '%s' as WAV file\n", fname_inp.c_str());
|
|
||||||
return 5;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (wav.channels != 1 && wav.channels != 2) {
|
|
||||||
fprintf(stderr, "error: WAV file '%s' must be mono or stereo\n", fname_inp.c_str());
|
|
||||||
return 6;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (params.diarize && wav.channels != 2 && params.no_timestamps == false) {
|
|
||||||
fprintf(stderr, "error: WAV file '%s' must be stereo for diarization and timestamps have to be enabled\n", fname_inp.c_str());
|
|
||||||
return 6;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (wav.sampleRate != WHISPER_SAMPLE_RATE) {
|
|
||||||
fprintf(stderr, "error: WAV file '%s' must be %i kHz\n", fname_inp.c_str(), WHISPER_SAMPLE_RATE/1000);
|
|
||||||
return 8;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (wav.bitsPerSample != 16) {
|
|
||||||
fprintf(stderr, "error: WAV file '%s' must be 16-bit\n", fname_inp.c_str());
|
|
||||||
return 9;
|
|
||||||
}
|
|
||||||
|
|
||||||
const uint64_t n = wav_data.empty() ? wav.totalPCMFrameCount : wav_data.size()/(wav.channels*wav.bitsPerSample/8);
|
|
||||||
|
|
||||||
std::vector<int16_t> pcm16;
|
|
||||||
pcm16.resize(n*wav.channels);
|
|
||||||
drwav_read_pcm_frames_s16(&wav, n, pcm16.data());
|
|
||||||
drwav_uninit(&wav);
|
|
||||||
|
|
||||||
// convert to mono, float
|
|
||||||
pcmf32.resize(n);
|
|
||||||
if (wav.channels == 1) {
|
|
||||||
for (uint64_t i = 0; i < n; i++) {
|
|
||||||
pcmf32[i] = float(pcm16[i])/32768.0f;
|
|
||||||
}
|
|
||||||
} else {
|
|
||||||
for (uint64_t i = 0; i < n; i++) {
|
|
||||||
pcmf32[i] = float(pcm16[2*i] + pcm16[2*i + 1])/65536.0f;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
if (params.diarize) {
|
|
||||||
// convert to stereo, float
|
|
||||||
pcmf32s.resize(2);
|
|
||||||
|
|
||||||
pcmf32s[0].resize(n);
|
|
||||||
pcmf32s[1].resize(n);
|
|
||||||
for (uint64_t i = 0; i < n; i++) {
|
|
||||||
pcmf32s[0][i] = float(pcm16[2*i])/32768.0f;
|
|
||||||
pcmf32s[1][i] = float(pcm16[2*i + 1])/32768.0f;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// print system information
|
|
||||||
{
|
|
||||||
fprintf(stderr, "\n");
|
|
||||||
fprintf(stderr, "system_info: n_threads = %d / %d | %s\n",
|
|
||||||
params.n_threads*params.n_processors, std::thread::hardware_concurrency(), whisper_print_system_info());
|
|
||||||
}
|
|
||||||
|
|
||||||
// print some info about the processing
|
|
||||||
{
|
|
||||||
fprintf(stderr, "\n");
|
|
||||||
if (!whisper_is_multilingual(ctx)) {
|
|
||||||
if (params.language != "en" || params.translate) {
|
|
||||||
params.language = "en";
|
|
||||||
params.translate = false;
|
|
||||||
fprintf(stderr, "%s: WARNING: model is not multilingual, ignoring language and translation options\n", __func__);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
fprintf(stderr, "%s: processing '%s' (%d samples, %.1f sec), %d threads, %d processors, lang = %s, task = %s, timestamps = %d ...\n",
|
|
||||||
__func__, fname_inp.c_str(), int(pcmf32.size()), float(pcmf32.size())/WHISPER_SAMPLE_RATE,
|
|
||||||
params.n_threads, params.n_processors,
|
|
||||||
params.language.c_str(),
|
|
||||||
params.translate ? "translate" : "transcribe",
|
|
||||||
params.no_timestamps ? 0 : 1);
|
|
||||||
|
|
||||||
fprintf(stderr, "\n");
|
|
||||||
}
|
|
||||||
|
|
||||||
// run the inference
|
|
||||||
{
|
|
||||||
whisper_full_params wparams = whisper_full_default_params(WHISPER_SAMPLING_GREEDY);
|
|
||||||
|
|
||||||
wparams.strategy = params.beam_size > 1 ? WHISPER_SAMPLING_BEAM_SEARCH : WHISPER_SAMPLING_GREEDY;
|
|
||||||
|
|
||||||
wparams.print_realtime = false;
|
|
||||||
wparams.print_progress = params.print_progress;
|
|
||||||
wparams.print_timestamps = !params.no_timestamps;
|
|
||||||
wparams.print_special = params.print_special;
|
|
||||||
wparams.translate = params.translate;
|
|
||||||
wparams.language = params.language.c_str();
|
|
||||||
wparams.n_threads = params.n_threads;
|
|
||||||
wparams.n_max_text_ctx = params.max_context >= 0 ? params.max_context : wparams.n_max_text_ctx;
|
|
||||||
wparams.offset_ms = params.offset_t_ms;
|
|
||||||
wparams.duration_ms = params.duration_ms;
|
|
||||||
|
|
||||||
wparams.token_timestamps = params.output_wts || params.max_len > 0;
|
|
||||||
wparams.thold_pt = params.word_thold;
|
|
||||||
wparams.entropy_thold = params.entropy_thold;
|
|
||||||
wparams.logprob_thold = params.logprob_thold;
|
|
||||||
wparams.max_len = params.output_wts && params.max_len == 0 ? 60 : params.max_len;
|
|
||||||
|
|
||||||
wparams.speed_up = params.speed_up;
|
|
||||||
|
|
||||||
wparams.greedy.best_of = params.best_of;
|
|
||||||
wparams.beam_search.beam_size = params.beam_size;
|
|
||||||
|
|
||||||
wparams.prompt_tokens = prompt_tokens.empty() ? nullptr : prompt_tokens.data();
|
|
||||||
wparams.prompt_n_tokens = prompt_tokens.empty() ? 0 : prompt_tokens.size();
|
|
||||||
|
|
||||||
whisper_print_user_data user_data = { ¶ms, &pcmf32s };
|
|
||||||
|
|
||||||
// this callback is called on each new segment
|
|
||||||
if (!wparams.print_realtime) {
|
|
||||||
wparams.new_segment_callback = whisper_print_segment_callback;
|
|
||||||
wparams.new_segment_callback_user_data = &user_data;
|
|
||||||
}
|
|
||||||
|
|
||||||
// example for abort mechanism
|
|
||||||
// in this example, we do not abort the processing, but we could if the flag is set to true
|
|
||||||
// the callback is called before every encoder run - if it returns false, the processing is aborted
|
|
||||||
{
|
|
||||||
static bool is_aborted = false; // NOTE: this should be atomic to avoid data race
|
|
||||||
|
|
||||||
wparams.encoder_begin_callback = [](struct whisper_context * /*ctx*/, void * user_data) {
|
|
||||||
bool is_aborted = *(bool*)user_data;
|
|
||||||
return !is_aborted;
|
|
||||||
};
|
|
||||||
wparams.encoder_begin_callback_user_data = &is_aborted;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (whisper_full_parallel(ctx, wparams, pcmf32.data(), pcmf32.size(), params.n_processors) != 0) {
|
|
||||||
fprintf(stderr, "failed to process audio\n");
|
|
||||||
return 10;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
const int n_segments = whisper_full_n_segments(ctx);
|
|
||||||
result.resize(n_segments);
|
|
||||||
for (int i = 0; i < n_segments; ++i) {
|
|
||||||
const char * text = whisper_full_get_segment_text(ctx, i);
|
|
||||||
const int64_t t0 = whisper_full_get_segment_t0(ctx, i);
|
|
||||||
const int64_t t1 = whisper_full_get_segment_t1(ctx, i);
|
|
||||||
|
|
||||||
result[i].emplace_back(to_timestamp(t0, true));
|
|
||||||
result[i].emplace_back(to_timestamp(t1, true));
|
|
||||||
result[i].emplace_back(text);
|
|
||||||
}
|
|
||||||
|
|
||||||
whisper_print_timings(ctx);
|
|
||||||
whisper_free(ctx);
|
|
||||||
|
|
||||||
return 0;
|
|
||||||
}
|
|
||||||
|
|
||||||
Napi::Object whisper(const Napi::CallbackInfo& info) {
|
|
||||||
Napi::Env env = info.Env();
|
|
||||||
if (info.Length() <= 0 || !info[0].IsObject()) {
|
|
||||||
Napi::TypeError::New(env, "object expected").ThrowAsJavaScriptException();
|
|
||||||
}
|
|
||||||
whisper_params params;
|
|
||||||
std::vector<std::vector<std::string>> result;
|
|
||||||
|
|
||||||
Napi::Object whisper_params = info[0].As<Napi::Object>();
|
|
||||||
std::string language = whisper_params.Get("language").As<Napi::String>();
|
|
||||||
std::string model = whisper_params.Get("model").As<Napi::String>();
|
|
||||||
std::string input = whisper_params.Get("fname_inp").As<Napi::String>();
|
|
||||||
|
|
||||||
params.language = language;
|
|
||||||
params.model = model;
|
|
||||||
params.fname_inp.emplace_back(input);
|
|
||||||
|
|
||||||
// run model
|
|
||||||
run(params, result);
|
|
||||||
|
|
||||||
fprintf(stderr, "RESULT:\n");
|
|
||||||
for (auto sentence:result) {
|
|
||||||
fprintf(stderr, "t0: %s, t1: %s, content: %s \n",
|
|
||||||
sentence[0].c_str(), sentence[1].c_str(), sentence[2].c_str());
|
|
||||||
}
|
|
||||||
|
|
||||||
Napi::Object res = Napi::Array::New(env, result.size());
|
|
||||||
for (uint64_t i = 0; i < result.size(); ++i) {
|
|
||||||
Napi::Object tmp = Napi::Array::New(env, 3);
|
|
||||||
for (uint64_t j = 0; j < 3; ++j) {
|
|
||||||
tmp[j] = Napi::String::New(env, result[i][j]);
|
|
||||||
}
|
|
||||||
res[i] = tmp;
|
|
||||||
}
|
|
||||||
|
|
||||||
return res;
|
|
||||||
}
|
|
||||||
|
|
||||||
|
|
||||||
Napi::Object Init(Napi::Env env, Napi::Object exports) {
|
|
||||||
exports.Set(
|
|
||||||
Napi::String::New(env, "whisper"),
|
|
||||||
Napi::Function::New(env, whisper)
|
|
||||||
);
|
|
||||||
return exports;
|
|
||||||
}
|
|
||||||
|
|
||||||
NODE_API_MODULE(whisper, Init);
|
|
||||||
@ -1,27 +0,0 @@
|
|||||||
const path = require('path');
|
|
||||||
const { whisper } = require(path.join(__dirname, '../../build/Release/whisper-addon'));
|
|
||||||
|
|
||||||
const whisperParams = {
|
|
||||||
language: 'en',
|
|
||||||
model: path.join(__dirname, '../../models/ggml-base.en.bin'),
|
|
||||||
fname_inp: '',
|
|
||||||
};
|
|
||||||
|
|
||||||
const arguments = process.argv.slice(2);
|
|
||||||
const params = Object.fromEntries(
|
|
||||||
arguments.reduce((pre, item) => {
|
|
||||||
if (item.startsWith("--")) {
|
|
||||||
return [...pre, item.slice(2).split("=")];
|
|
||||||
}
|
|
||||||
return pre;
|
|
||||||
}, []),
|
|
||||||
);
|
|
||||||
|
|
||||||
for (const key in params) {
|
|
||||||
if (whisperParams.hasOwnProperty(key)) {
|
|
||||||
whisperParams[key] = params[key];
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
console.log('whisperParams =', whisperParams);
|
|
||||||
console.log(whisper(whisperParams));
|
|
||||||
@ -1,16 +0,0 @@
|
|||||||
{
|
|
||||||
"name": "whisper-addon",
|
|
||||||
"version": "0.0.0",
|
|
||||||
"description": "",
|
|
||||||
"main": "index.js",
|
|
||||||
"author": "Qanhe Chen",
|
|
||||||
"license": "MIT",
|
|
||||||
"scripts": {
|
|
||||||
"test": "jest"
|
|
||||||
},
|
|
||||||
"devDependencies": {
|
|
||||||
"cmake-js": "^7.1.1",
|
|
||||||
"jest": "^29.4.0",
|
|
||||||
"node-addon-api": "^5.0.0"
|
|
||||||
}
|
|
||||||
}
|
|
||||||
@ -1,10 +0,0 @@
|
|||||||
if (WHISPER_SUPPORT_SDL2)
|
|
||||||
# chess
|
|
||||||
set(TARGET chess)
|
|
||||||
add_executable(${TARGET} chess.cpp)
|
|
||||||
|
|
||||||
include(DefaultTargetOptions)
|
|
||||||
|
|
||||||
target_include_directories(${TARGET} PRIVATE ${SDL2_INCLUDE_DIRS})
|
|
||||||
target_link_libraries(${TARGET} PRIVATE common whisper ${SDL2_LIBRARIES} ${CMAKE_THREAD_LIBS_INIT})
|
|
||||||
endif ()
|
|
||||||
@ -1,634 +0,0 @@
|
|||||||
// Input chess moves via voice
|
|
||||||
//
|
|
||||||
|
|
||||||
#include "common.h"
|
|
||||||
#include "whisper.h"
|
|
||||||
|
|
||||||
#include <SDL.h>
|
|
||||||
#include <SDL_audio.h>
|
|
||||||
|
|
||||||
#include <atomic>
|
|
||||||
#include <cassert>
|
|
||||||
#include <cstdio>
|
|
||||||
#include <string>
|
|
||||||
#include <thread>
|
|
||||||
#include <vector>
|
|
||||||
#include <fstream>
|
|
||||||
#include <mutex>
|
|
||||||
|
|
||||||
// 500 -> 00:05.000
|
|
||||||
// 6000 -> 01:00.000
|
|
||||||
std::string to_timestamp(int64_t t) {
|
|
||||||
int64_t sec = t/100;
|
|
||||||
int64_t msec = t - sec*100;
|
|
||||||
int64_t min = sec/60;
|
|
||||||
sec = sec - min*60;
|
|
||||||
|
|
||||||
char buf[32];
|
|
||||||
snprintf(buf, sizeof(buf), "%02d:%02d.%03d", (int) min, (int) sec, (int) msec);
|
|
||||||
|
|
||||||
return std::string(buf);
|
|
||||||
}
|
|
||||||
|
|
||||||
// command-line parameters
|
|
||||||
struct whisper_params {
|
|
||||||
int32_t n_threads = std::min(4, (int32_t) std::thread::hardware_concurrency());
|
|
||||||
int32_t step_ms = 3000;
|
|
||||||
int32_t length_ms = 10000;
|
|
||||||
int32_t keep_ms = 200;
|
|
||||||
int32_t capture_id = -1;
|
|
||||||
int32_t max_tokens = 32;
|
|
||||||
int32_t audio_ctx = 0;
|
|
||||||
|
|
||||||
float vad_thold = 0.6f;
|
|
||||||
float freq_thold = 100.0f;
|
|
||||||
|
|
||||||
bool translate = false;
|
|
||||||
bool print_special = false;
|
|
||||||
bool no_context = true;
|
|
||||||
bool no_timestamps = false;
|
|
||||||
|
|
||||||
std::string language = "en";
|
|
||||||
std::string model = "models/ggml-base.en.bin";
|
|
||||||
std::string fname_inp;
|
|
||||||
};
|
|
||||||
|
|
||||||
void whisper_print_usage(int argc, char ** argv, const whisper_params & params);
|
|
||||||
|
|
||||||
bool whisper_params_parse(int argc, char ** argv, whisper_params & params) {
|
|
||||||
for (int i = 1; i < argc; i++) {
|
|
||||||
std::string arg = argv[i];
|
|
||||||
|
|
||||||
if (arg == "-h" || arg == "--help") {
|
|
||||||
whisper_print_usage(argc, argv, params);
|
|
||||||
exit(0);
|
|
||||||
}
|
|
||||||
else if (arg == "-t" || arg == "--threads") { params.n_threads = std::stoi(argv[++i]); }
|
|
||||||
else if ( arg == "--step") { params.step_ms = std::stoi(argv[++i]); }
|
|
||||||
else if ( arg == "--length") { params.length_ms = std::stoi(argv[++i]); }
|
|
||||||
else if ( arg == "--keep") { params.keep_ms = std::stoi(argv[++i]); }
|
|
||||||
else if (arg == "-c" || arg == "--capture") { params.capture_id = std::stoi(argv[++i]); }
|
|
||||||
else if (arg == "-mt" || arg == "--max-tokens") { params.max_tokens = std::stoi(argv[++i]); }
|
|
||||||
else if (arg == "-ac" || arg == "--audio-ctx") { params.audio_ctx = std::stoi(argv[++i]); }
|
|
||||||
else if (arg == "-vth" || arg == "--vad-thold") { params.vad_thold = std::stof(argv[++i]); }
|
|
||||||
else if (arg == "-fth" || arg == "--freq-thold") { params.freq_thold = std::stof(argv[++i]); }
|
|
||||||
else if (arg == "-tr" || arg == "--translate") { params.translate = true; }
|
|
||||||
else if (arg == "-ps" || arg == "--print-special") { params.print_special = true; }
|
|
||||||
else if (arg == "-kc" || arg == "--keep-context") { params.no_context = false; }
|
|
||||||
else if (arg == "-l" || arg == "--language") { params.language = argv[++i]; }
|
|
||||||
else if (arg == "-m" || arg == "--model") { params.model = argv[++i]; }
|
|
||||||
else if (arg == "-f" || arg == "--file") { params.fname_inp = argv[++i]; }
|
|
||||||
else {
|
|
||||||
fprintf(stderr, "error: unknown argument: %s\n", arg.c_str());
|
|
||||||
whisper_print_usage(argc, argv, params);
|
|
||||||
exit(0);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
return true;
|
|
||||||
}
|
|
||||||
|
|
||||||
void whisper_print_usage(int /*argc*/, char ** argv, const whisper_params & params) {
|
|
||||||
fprintf(stderr, "\n");
|
|
||||||
fprintf(stderr, "usage: %s [options]\n", argv[0]);
|
|
||||||
fprintf(stderr, "\n");
|
|
||||||
fprintf(stderr, "options:\n");
|
|
||||||
fprintf(stderr, " -h, --help [default] show this help message and exit\n");
|
|
||||||
fprintf(stderr, " -t N, --threads N [%-7d] number of threads to use during computation\n", params.n_threads);
|
|
||||||
fprintf(stderr, " --step N [%-7d] audio step size in milliseconds\n", params.step_ms);
|
|
||||||
fprintf(stderr, " --length N [%-7d] audio length in milliseconds\n", params.length_ms);
|
|
||||||
fprintf(stderr, " --keep N [%-7d] audio to keep from previous step in ms\n", params.keep_ms);
|
|
||||||
fprintf(stderr, " -c ID, --capture ID [%-7d] capture device ID\n", params.capture_id);
|
|
||||||
fprintf(stderr, " -mt N, --max-tokens N [%-7d] maximum number of tokens per audio chunk\n", params.max_tokens);
|
|
||||||
fprintf(stderr, " -ac N, --audio-ctx N [%-7d] audio context size (0 - all)\n", params.audio_ctx);
|
|
||||||
fprintf(stderr, " -vth N, --vad-thold N [%-7.2f] voice activity detection threshold\n", params.vad_thold);
|
|
||||||
fprintf(stderr, " -fth N, --freq-thold N [%-7.2f] high-pass frequency cutoff\n", params.freq_thold);
|
|
||||||
fprintf(stderr, " -tr, --translate [%-7s] translate from source language to english\n", params.translate ? "true" : "false");
|
|
||||||
fprintf(stderr, " -ps, --print-special [%-7s] print special tokens\n", params.print_special ? "true" : "false");
|
|
||||||
fprintf(stderr, " -kc, --keep-context [%-7s] keep context between audio chunks\n", params.no_context ? "false" : "true");
|
|
||||||
fprintf(stderr, " -l LANG, --language LANG [%-7s] spoken language\n", params.language.c_str());
|
|
||||||
fprintf(stderr, " -m FNAME, --model FNAME [%-7s] model path\n", params.model.c_str());
|
|
||||||
fprintf(stderr, " -f FNAME, --file FNAME [%-7s] input WAV file path\n", params.fname_inp.c_str());
|
|
||||||
fprintf(stderr, "\n");
|
|
||||||
}
|
|
||||||
|
|
||||||
//
|
|
||||||
// SDL Audio capture
|
|
||||||
//
|
|
||||||
|
|
||||||
class audio_async {
|
|
||||||
public:
|
|
||||||
audio_async(int len_ms);
|
|
||||||
~audio_async();
|
|
||||||
|
|
||||||
bool init(int capture_id, int sample_rate);
|
|
||||||
|
|
||||||
// start capturing audio via the provided SDL callback
|
|
||||||
// keep last len_ms seconds of audio in a circular buffer
|
|
||||||
bool resume();
|
|
||||||
bool pause();
|
|
||||||
bool clear();
|
|
||||||
|
|
||||||
// callback to be called by SDL
|
|
||||||
void callback(uint8_t * stream, int len);
|
|
||||||
|
|
||||||
// get audio data from the circular buffer
|
|
||||||
void get(int ms, std::vector<float> & audio);
|
|
||||||
|
|
||||||
private:
|
|
||||||
SDL_AudioDeviceID m_dev_id_in = 0;
|
|
||||||
|
|
||||||
int m_len_ms = 0;
|
|
||||||
int m_sample_rate = 0;
|
|
||||||
|
|
||||||
std::atomic_bool m_running;
|
|
||||||
std::mutex m_mutex;
|
|
||||||
|
|
||||||
std::vector<float> m_audio;
|
|
||||||
std::vector<float> m_audio_new;
|
|
||||||
size_t m_audio_pos = 0;
|
|
||||||
size_t m_audio_len = 0;
|
|
||||||
};
|
|
||||||
|
|
||||||
audio_async::audio_async(int len_ms) {
|
|
||||||
m_len_ms = len_ms;
|
|
||||||
|
|
||||||
m_running = false;
|
|
||||||
}
|
|
||||||
|
|
||||||
audio_async::~audio_async() {
|
|
||||||
if (m_dev_id_in) {
|
|
||||||
SDL_CloseAudioDevice(m_dev_id_in);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
bool audio_async::init(int capture_id, int sample_rate) {
|
|
||||||
SDL_LogSetPriority(SDL_LOG_CATEGORY_APPLICATION, SDL_LOG_PRIORITY_INFO);
|
|
||||||
|
|
||||||
if (SDL_Init(SDL_INIT_AUDIO) < 0) {
|
|
||||||
SDL_LogError(SDL_LOG_CATEGORY_APPLICATION, "Couldn't initialize SDL: %s\n", SDL_GetError());
|
|
||||||
return false;
|
|
||||||
}
|
|
||||||
|
|
||||||
SDL_SetHintWithPriority(SDL_HINT_AUDIO_RESAMPLING_MODE, "medium", SDL_HINT_OVERRIDE);
|
|
||||||
|
|
||||||
{
|
|
||||||
int nDevices = SDL_GetNumAudioDevices(SDL_TRUE);
|
|
||||||
fprintf(stderr, "%s: found %d capture devices:\n", __func__, nDevices);
|
|
||||||
for (int i = 0; i < nDevices; i++) {
|
|
||||||
fprintf(stderr, "%s: - Capture device #%d: '%s'\n", __func__, i, SDL_GetAudioDeviceName(i, SDL_TRUE));
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
SDL_AudioSpec capture_spec_requested;
|
|
||||||
SDL_AudioSpec capture_spec_obtained;
|
|
||||||
|
|
||||||
SDL_zero(capture_spec_requested);
|
|
||||||
SDL_zero(capture_spec_obtained);
|
|
||||||
|
|
||||||
capture_spec_requested.freq = sample_rate;
|
|
||||||
capture_spec_requested.format = AUDIO_F32;
|
|
||||||
capture_spec_requested.channels = 1;
|
|
||||||
capture_spec_requested.samples = 1024;
|
|
||||||
capture_spec_requested.callback = [](void * userdata, uint8_t * stream, int len) {
|
|
||||||
audio_async * audio = (audio_async *) userdata;
|
|
||||||
audio->callback(stream, len);
|
|
||||||
};
|
|
||||||
capture_spec_requested.userdata = this;
|
|
||||||
|
|
||||||
if (capture_id >= 0) {
|
|
||||||
fprintf(stderr, "%s: attempt to open capture device %d : '%s' ...\n", __func__, capture_id, SDL_GetAudioDeviceName(capture_id, SDL_TRUE));
|
|
||||||
m_dev_id_in = SDL_OpenAudioDevice(SDL_GetAudioDeviceName(capture_id, SDL_TRUE), SDL_TRUE, &capture_spec_requested, &capture_spec_obtained, 0);
|
|
||||||
} else {
|
|
||||||
fprintf(stderr, "%s: attempt to open default capture device ...\n", __func__);
|
|
||||||
m_dev_id_in = SDL_OpenAudioDevice(nullptr, SDL_TRUE, &capture_spec_requested, &capture_spec_obtained, 0);
|
|
||||||
}
|
|
||||||
|
|
||||||
if (!m_dev_id_in) {
|
|
||||||
fprintf(stderr, "%s: couldn't open an audio device for capture: %s!\n", __func__, SDL_GetError());
|
|
||||||
m_dev_id_in = 0;
|
|
||||||
|
|
||||||
return false;
|
|
||||||
} else {
|
|
||||||
fprintf(stderr, "%s: obtained spec for input device (SDL Id = %d):\n", __func__, m_dev_id_in);
|
|
||||||
fprintf(stderr, "%s: - sample rate: %d\n", __func__, capture_spec_obtained.freq);
|
|
||||||
fprintf(stderr, "%s: - format: %d (required: %d)\n", __func__, capture_spec_obtained.format,
|
|
||||||
capture_spec_requested.format);
|
|
||||||
fprintf(stderr, "%s: - channels: %d (required: %d)\n", __func__, capture_spec_obtained.channels,
|
|
||||||
capture_spec_requested.channels);
|
|
||||||
fprintf(stderr, "%s: - samples per frame: %d\n", __func__, capture_spec_obtained.samples);
|
|
||||||
}
|
|
||||||
|
|
||||||
m_sample_rate = capture_spec_obtained.freq;
|
|
||||||
|
|
||||||
m_audio.resize((m_sample_rate*m_len_ms)/1000);
|
|
||||||
|
|
||||||
return true;
|
|
||||||
}
|
|
||||||
|
|
||||||
bool audio_async::resume() {
|
|
||||||
if (!m_dev_id_in) {
|
|
||||||
fprintf(stderr, "%s: no audio device to resume!\n", __func__);
|
|
||||||
return false;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (m_running) {
|
|
||||||
fprintf(stderr, "%s: already running!\n", __func__);
|
|
||||||
return false;
|
|
||||||
}
|
|
||||||
|
|
||||||
SDL_PauseAudioDevice(m_dev_id_in, 0);
|
|
||||||
|
|
||||||
m_running = true;
|
|
||||||
|
|
||||||
return true;
|
|
||||||
}
|
|
||||||
|
|
||||||
bool audio_async::pause() {
|
|
||||||
if (!m_dev_id_in) {
|
|
||||||
fprintf(stderr, "%s: no audio device to pause!\n", __func__);
|
|
||||||
return false;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (!m_running) {
|
|
||||||
fprintf(stderr, "%s: already paused!\n", __func__);
|
|
||||||
return false;
|
|
||||||
}
|
|
||||||
|
|
||||||
SDL_PauseAudioDevice(m_dev_id_in, 1);
|
|
||||||
|
|
||||||
m_running = false;
|
|
||||||
|
|
||||||
return true;
|
|
||||||
}
|
|
||||||
|
|
||||||
bool audio_async::clear() {
|
|
||||||
if (!m_dev_id_in) {
|
|
||||||
fprintf(stderr, "%s: no audio device to clear!\n", __func__);
|
|
||||||
return false;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (!m_running) {
|
|
||||||
fprintf(stderr, "%s: not running!\n", __func__);
|
|
||||||
return false;
|
|
||||||
}
|
|
||||||
|
|
||||||
{
|
|
||||||
std::lock_guard<std::mutex> lock(m_mutex);
|
|
||||||
|
|
||||||
m_audio_pos = 0;
|
|
||||||
m_audio_len = 0;
|
|
||||||
}
|
|
||||||
|
|
||||||
return true;
|
|
||||||
}
|
|
||||||
|
|
||||||
// callback to be called by SDL
|
|
||||||
void audio_async::callback(uint8_t * stream, int len) {
|
|
||||||
if (!m_running) {
|
|
||||||
return;
|
|
||||||
}
|
|
||||||
|
|
||||||
const size_t n_samples = len / sizeof(float);
|
|
||||||
|
|
||||||
m_audio_new.resize(n_samples);
|
|
||||||
memcpy(m_audio_new.data(), stream, n_samples * sizeof(float));
|
|
||||||
|
|
||||||
//fprintf(stderr, "%s: %zu samples, pos %zu, len %zu\n", __func__, n_samples, m_audio_pos, m_audio_len);
|
|
||||||
|
|
||||||
{
|
|
||||||
std::lock_guard<std::mutex> lock(m_mutex);
|
|
||||||
|
|
||||||
if (m_audio_pos + n_samples > m_audio.size()) {
|
|
||||||
const size_t n0 = m_audio.size() - m_audio_pos;
|
|
||||||
|
|
||||||
memcpy(&m_audio[m_audio_pos], stream, n0 * sizeof(float));
|
|
||||||
memcpy(&m_audio[0], &stream[n0], (n_samples - n0) * sizeof(float));
|
|
||||||
|
|
||||||
m_audio_pos = (m_audio_pos + n_samples) % m_audio.size();
|
|
||||||
m_audio_len = m_audio.size();
|
|
||||||
} else {
|
|
||||||
memcpy(&m_audio[m_audio_pos], stream, n_samples * sizeof(float));
|
|
||||||
|
|
||||||
m_audio_pos = (m_audio_pos + n_samples) % m_audio.size();
|
|
||||||
m_audio_len = std::min(m_audio_len + n_samples, m_audio.size());
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
void audio_async::get(int ms, std::vector<float> & result) {
|
|
||||||
if (!m_dev_id_in) {
|
|
||||||
fprintf(stderr, "%s: no audio device to get audio from!\n", __func__);
|
|
||||||
return;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (!m_running) {
|
|
||||||
fprintf(stderr, "%s: not running!\n", __func__);
|
|
||||||
return;
|
|
||||||
}
|
|
||||||
|
|
||||||
result.clear();
|
|
||||||
|
|
||||||
{
|
|
||||||
std::lock_guard<std::mutex> lock(m_mutex);
|
|
||||||
|
|
||||||
if (ms <= 0) {
|
|
||||||
ms = m_len_ms;
|
|
||||||
}
|
|
||||||
|
|
||||||
size_t n_samples = (m_sample_rate * ms) / 1000;
|
|
||||||
if (n_samples > m_audio_len) {
|
|
||||||
n_samples = m_audio_len;
|
|
||||||
}
|
|
||||||
|
|
||||||
result.resize(n_samples);
|
|
||||||
|
|
||||||
int s0 = m_audio_pos - n_samples;
|
|
||||||
if (s0 < 0) {
|
|
||||||
s0 += m_audio.size();
|
|
||||||
}
|
|
||||||
|
|
||||||
if (s0 + n_samples > m_audio.size()) {
|
|
||||||
const size_t n0 = m_audio.size() - s0;
|
|
||||||
|
|
||||||
memcpy(result.data(), &m_audio[s0], n0 * sizeof(float));
|
|
||||||
memcpy(&result[n0], &m_audio[0], (n_samples - n0) * sizeof(float));
|
|
||||||
} else {
|
|
||||||
memcpy(result.data(), &m_audio[s0], n_samples * sizeof(float));
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
///////////////////////////
|
|
||||||
|
|
||||||
int main(int argc, char ** argv) {
|
|
||||||
whisper_params params;
|
|
||||||
|
|
||||||
if (whisper_params_parse(argc, argv, params) == false) {
|
|
||||||
return 1;
|
|
||||||
}
|
|
||||||
|
|
||||||
params.keep_ms = std::min(params.keep_ms, params.step_ms);
|
|
||||||
params.length_ms = std::max(params.length_ms, params.step_ms);
|
|
||||||
|
|
||||||
const int n_samples_step = (1e-3*params.step_ms )*WHISPER_SAMPLE_RATE;
|
|
||||||
const int n_samples_len = (1e-3*params.length_ms)*WHISPER_SAMPLE_RATE;
|
|
||||||
const int n_samples_keep = (1e-3*params.keep_ms )*WHISPER_SAMPLE_RATE;
|
|
||||||
const int n_samples_30s = (1e-3*30000.0 )*WHISPER_SAMPLE_RATE;
|
|
||||||
|
|
||||||
const bool use_vad = n_samples_step <= 0; // sliding window mode uses VAD
|
|
||||||
|
|
||||||
const int n_new_line = !use_vad ? std::max(1, params.length_ms / params.step_ms - 1) : 1; // number of steps to print new line
|
|
||||||
|
|
||||||
params.no_timestamps = !use_vad;
|
|
||||||
params.no_context |= use_vad;
|
|
||||||
params.max_tokens = 0;
|
|
||||||
|
|
||||||
// init audio
|
|
||||||
|
|
||||||
audio_async audio(params.length_ms);
|
|
||||||
if (!audio.init(params.capture_id, WHISPER_SAMPLE_RATE)) {
|
|
||||||
fprintf(stderr, "%s: audio.init() failed!\n", __func__);
|
|
||||||
return 1;
|
|
||||||
}
|
|
||||||
|
|
||||||
audio.resume();
|
|
||||||
|
|
||||||
// whisper init
|
|
||||||
|
|
||||||
if (whisper_lang_id(params.language.c_str()) == -1) {
|
|
||||||
fprintf(stderr, "error: unknown language '%s'\n", params.language.c_str());
|
|
||||||
whisper_print_usage(argc, argv, params);
|
|
||||||
exit(0);
|
|
||||||
}
|
|
||||||
|
|
||||||
struct whisper_context * ctx = whisper_init_from_file(params.model.c_str());
|
|
||||||
|
|
||||||
std::vector<float> pcmf32 (n_samples_30s, 0.0f);
|
|
||||||
std::vector<float> pcmf32_old;
|
|
||||||
std::vector<float> pcmf32_new(n_samples_30s, 0.0f);
|
|
||||||
|
|
||||||
std::vector<whisper_token> prompt_tokens;
|
|
||||||
|
|
||||||
// print some info about the processing
|
|
||||||
{
|
|
||||||
fprintf(stderr, "\n");
|
|
||||||
if (!whisper_is_multilingual(ctx)) {
|
|
||||||
if (params.language != "en" || params.translate) {
|
|
||||||
params.language = "en";
|
|
||||||
params.translate = false;
|
|
||||||
fprintf(stderr, "%s: WARNING: model is not multilingual, ignoring language and translation options\n", __func__);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
fprintf(stderr, "%s: processing %d samples (step = %.1f sec / len = %.1f sec / keep = %.1f sec), %d threads, lang = %s, task = %s, timestamps = %d ...\n",
|
|
||||||
__func__,
|
|
||||||
n_samples_step,
|
|
||||||
float(n_samples_step)/WHISPER_SAMPLE_RATE,
|
|
||||||
float(n_samples_len )/WHISPER_SAMPLE_RATE,
|
|
||||||
float(n_samples_keep)/WHISPER_SAMPLE_RATE,
|
|
||||||
params.n_threads,
|
|
||||||
params.language.c_str(),
|
|
||||||
params.translate ? "translate" : "transcribe",
|
|
||||||
params.no_timestamps ? 0 : 1);
|
|
||||||
|
|
||||||
if (!use_vad) {
|
|
||||||
fprintf(stderr, "%s: n_new_line = %d, no_context = %d\n", __func__, n_new_line, params.no_context);
|
|
||||||
} else {
|
|
||||||
fprintf(stderr, "%s: using VAD, will transcribe on speech activity\n", __func__);
|
|
||||||
}
|
|
||||||
|
|
||||||
fprintf(stderr, "\n");
|
|
||||||
}
|
|
||||||
|
|
||||||
int n_iter = 0;
|
|
||||||
|
|
||||||
bool is_running = true;
|
|
||||||
|
|
||||||
printf("[Start speaking]");
|
|
||||||
fflush(stdout);
|
|
||||||
|
|
||||||
auto t_last = std::chrono::high_resolution_clock::now();
|
|
||||||
const auto t_start = t_last;
|
|
||||||
|
|
||||||
// main audio loop
|
|
||||||
while (is_running) {
|
|
||||||
// handle Ctrl + C
|
|
||||||
{
|
|
||||||
SDL_Event event;
|
|
||||||
while (SDL_PollEvent(&event)) {
|
|
||||||
switch (event.type) {
|
|
||||||
case SDL_QUIT:
|
|
||||||
{
|
|
||||||
is_running = false;
|
|
||||||
} break;
|
|
||||||
default:
|
|
||||||
break;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
if (!is_running) {
|
|
||||||
break;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
if (!is_running) {
|
|
||||||
break;
|
|
||||||
}
|
|
||||||
|
|
||||||
// process new audio
|
|
||||||
|
|
||||||
if (!use_vad) {
|
|
||||||
while (true) {
|
|
||||||
audio.get(params.step_ms, pcmf32_new);
|
|
||||||
|
|
||||||
if ((int) pcmf32_new.size() > 2*n_samples_step) {
|
|
||||||
fprintf(stderr, "\n\n%s: WARNING: cannot process audio fast enough, dropping audio ...\n\n", __func__);
|
|
||||||
audio.clear();
|
|
||||||
continue;
|
|
||||||
}
|
|
||||||
|
|
||||||
if ((int) pcmf32_new.size() >= n_samples_step) {
|
|
||||||
audio.clear();
|
|
||||||
break;
|
|
||||||
}
|
|
||||||
|
|
||||||
SDL_Delay(1);
|
|
||||||
}
|
|
||||||
|
|
||||||
const int n_samples_new = pcmf32_new.size();
|
|
||||||
|
|
||||||
// take up to params.length_ms audio from previous iteration
|
|
||||||
const int n_samples_take = std::min((int) pcmf32_old.size(), std::max(0, n_samples_keep + n_samples_len - n_samples_new));
|
|
||||||
|
|
||||||
//printf("processing: take = %d, new = %d, old = %d\n", n_samples_take, n_samples_new, (int) pcmf32_old.size());
|
|
||||||
|
|
||||||
pcmf32.resize(n_samples_new + n_samples_take);
|
|
||||||
|
|
||||||
for (int i = 0; i < n_samples_take; i++) {
|
|
||||||
pcmf32[i] = pcmf32_old[pcmf32_old.size() - n_samples_take + i];
|
|
||||||
}
|
|
||||||
|
|
||||||
memcpy(pcmf32.data() + n_samples_take, pcmf32_new.data(), n_samples_new*sizeof(float));
|
|
||||||
|
|
||||||
pcmf32_old = pcmf32;
|
|
||||||
} else {
|
|
||||||
const auto t_now = std::chrono::high_resolution_clock::now();
|
|
||||||
const auto t_diff = std::chrono::duration_cast<std::chrono::milliseconds>(t_now - t_last).count();
|
|
||||||
|
|
||||||
if (t_diff < 2000) {
|
|
||||||
std::this_thread::sleep_for(std::chrono::milliseconds(100));
|
|
||||||
|
|
||||||
continue;
|
|
||||||
}
|
|
||||||
|
|
||||||
audio.get(2000, pcmf32_new);
|
|
||||||
|
|
||||||
if (vad_simple(pcmf32_new, WHISPER_SAMPLE_RATE, 1000, params.vad_thold, params.freq_thold, false)) {
|
|
||||||
audio.get(params.length_ms, pcmf32);
|
|
||||||
} else {
|
|
||||||
std::this_thread::sleep_for(std::chrono::milliseconds(100));
|
|
||||||
|
|
||||||
continue;
|
|
||||||
}
|
|
||||||
|
|
||||||
t_last = t_now;
|
|
||||||
}
|
|
||||||
|
|
||||||
// run the inference
|
|
||||||
{
|
|
||||||
whisper_full_params wparams = whisper_full_default_params(WHISPER_SAMPLING_GREEDY);
|
|
||||||
|
|
||||||
wparams.print_progress = false;
|
|
||||||
wparams.print_special = params.print_special;
|
|
||||||
wparams.print_realtime = false;
|
|
||||||
wparams.print_timestamps = !params.no_timestamps;
|
|
||||||
wparams.translate = params.translate;
|
|
||||||
wparams.no_context = true;
|
|
||||||
wparams.single_segment = !use_vad;
|
|
||||||
wparams.max_tokens = params.max_tokens;
|
|
||||||
wparams.language = params.language.c_str();
|
|
||||||
wparams.n_threads = params.n_threads;
|
|
||||||
|
|
||||||
wparams.audio_ctx = params.audio_ctx;
|
|
||||||
|
|
||||||
// disable temperature fallback
|
|
||||||
wparams.temperature_inc = -1.0f;
|
|
||||||
|
|
||||||
wparams.prompt_tokens = params.no_context ? nullptr : prompt_tokens.data();
|
|
||||||
wparams.prompt_n_tokens = params.no_context ? 0 : prompt_tokens.size();
|
|
||||||
|
|
||||||
if (whisper_full(ctx, wparams, pcmf32.data(), pcmf32.size()) != 0) {
|
|
||||||
fprintf(stderr, "%s: failed to process audio\n", argv[0]);
|
|
||||||
return 6;
|
|
||||||
}
|
|
||||||
|
|
||||||
// print result;
|
|
||||||
{
|
|
||||||
if (!use_vad) {
|
|
||||||
printf("\33[2K\r");
|
|
||||||
|
|
||||||
// print long empty line to clear the previous line
|
|
||||||
printf("%s", std::string(100, ' ').c_str());
|
|
||||||
|
|
||||||
printf("\33[2K\r");
|
|
||||||
} else {
|
|
||||||
const int64_t t1 = (t_last - t_start).count()/1000000;
|
|
||||||
const int64_t t0 = std::max(0.0, t1 - pcmf32.size()*1000.0/WHISPER_SAMPLE_RATE);
|
|
||||||
|
|
||||||
printf("\n");
|
|
||||||
printf("### Transcription %d START | t0 = %d ms | t1 = %d ms\n", n_iter, (int) t0, (int) t1);
|
|
||||||
printf("\n");
|
|
||||||
}
|
|
||||||
|
|
||||||
const int n_segments = whisper_full_n_segments(ctx);
|
|
||||||
for (int i = 0; i < n_segments; ++i) {
|
|
||||||
const char * text = whisper_full_get_segment_text(ctx, i);
|
|
||||||
|
|
||||||
if (params.no_timestamps) {
|
|
||||||
printf("%s", text);
|
|
||||||
fflush(stdout);
|
|
||||||
} else {
|
|
||||||
const int64_t t0 = whisper_full_get_segment_t0(ctx, i);
|
|
||||||
const int64_t t1 = whisper_full_get_segment_t1(ctx, i);
|
|
||||||
|
|
||||||
printf ("[%s --> %s] %s\n", to_timestamp(t0).c_str(), to_timestamp(t1).c_str(), text);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
if (use_vad){
|
|
||||||
printf("\n");
|
|
||||||
printf("### Transcription %d END\n", n_iter);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
++n_iter;
|
|
||||||
|
|
||||||
if (!use_vad && (n_iter % n_new_line) == 0) {
|
|
||||||
printf("\n");
|
|
||||||
|
|
||||||
// keep part of the audio for next iteration to try to mitigate word boundary issues
|
|
||||||
pcmf32_old = std::vector<float>(pcmf32.end() - n_samples_keep, pcmf32.end());
|
|
||||||
|
|
||||||
// Add tokens of the last full length segment as the prompt
|
|
||||||
if (!params.no_context) {
|
|
||||||
prompt_tokens.clear();
|
|
||||||
|
|
||||||
const int n_segments = whisper_full_n_segments(ctx);
|
|
||||||
for (int i = 0; i < n_segments; ++i) {
|
|
||||||
const int token_count = whisper_full_n_tokens(ctx, i);
|
|
||||||
for (int j = 0; j < token_count; ++j) {
|
|
||||||
prompt_tokens.push_back(whisper_full_get_token_id(ctx, i, j));
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
audio.pause();
|
|
||||||
|
|
||||||
whisper_print_timings(ctx);
|
|
||||||
whisper_free(ctx);
|
|
||||||
|
|
||||||
return 0;
|
|
||||||
}
|
|
||||||
@ -8,7 +8,7 @@ function convertTypedArray(src, type) {
|
|||||||
|
|
||||||
var printTextarea = (function() {
|
var printTextarea = (function() {
|
||||||
var element = document.getElementById('output');
|
var element = document.getElementById('output');
|
||||||
if (element) element.value = ''; // clear browser cache
|
if (element) element.alue = ''; // clear browser cache
|
||||||
return function(text) {
|
return function(text) {
|
||||||
if (arguments.length > 1) text = Array.prototype.slice.call(arguments).join(' ');
|
if (arguments.length > 1) text = Array.prototype.slice.call(arguments).join(' ');
|
||||||
console.log(text);
|
console.log(text);
|
||||||
@ -88,15 +88,11 @@ async function fetchRemote(url, cbProgress, cbPrint) {
|
|||||||
// - check if the data is already in the IndexedDB
|
// - check if the data is already in the IndexedDB
|
||||||
// - if not, fetch it from the remote URL and store it in the IndexedDB
|
// - if not, fetch it from the remote URL and store it in the IndexedDB
|
||||||
function loadRemote(url, dst, size_mb, cbProgress, cbReady, cbCancel, cbPrint) {
|
function loadRemote(url, dst, size_mb, cbProgress, cbReady, cbCancel, cbPrint) {
|
||||||
if (!navigator.storage || !navigator.storage.estimate) {
|
|
||||||
cbPrint('loadRemote: navigator.storage.estimate() is not supported');
|
|
||||||
} else {
|
|
||||||
// query the storage quota and print it
|
// query the storage quota and print it
|
||||||
navigator.storage.estimate().then(function (estimate) {
|
navigator.storage.estimate().then(function (estimate) {
|
||||||
cbPrint('loadRemote: storage quota: ' + estimate.quota + ' bytes');
|
cbPrint('loadRemote: storage quota: ' + estimate.quota + ' bytes');
|
||||||
cbPrint('loadRemote: storage usage: ' + estimate.usage + ' bytes');
|
cbPrint('loadRemote: storage usage: ' + estimate.usage + ' bytes');
|
||||||
});
|
});
|
||||||
}
|
|
||||||
|
|
||||||
// check if the data is already in the IndexedDB
|
// check if the data is already in the IndexedDB
|
||||||
var rq = indexedDB.open(dbName, dbVersion);
|
var rq = indexedDB.open(dbName, dbVersion);
|
||||||
|
|||||||
@ -100,7 +100,7 @@ while [ $running -eq 1 ]; do
|
|||||||
err=$(cat /tmp/whisper-live.err | wc -l)
|
err=$(cat /tmp/whisper-live.err | wc -l)
|
||||||
done
|
done
|
||||||
|
|
||||||
./main -t 8 -m ./models/ggml-${model}.bin -f /tmp/whisper-live.wav --no-timestamps -otxt 2> /tmp/whispererr | tail -n 1
|
./main -t 8 -m ./models/ggml-base.en.bin -f /tmp/whisper-live.wav --no-timestamps -otxt 2> /tmp/whispererr | tail -n 1
|
||||||
|
|
||||||
while [ $SECONDS -lt $((($i+1)*$step_s)) ]; do
|
while [ $SECONDS -lt $((($i+1)*$step_s)) ]; do
|
||||||
sleep 1
|
sleep 1
|
||||||
|
|||||||
@ -17,27 +17,17 @@ options:
|
|||||||
-d N, --duration N [0 ] duration of audio to process in milliseconds
|
-d N, --duration N [0 ] duration of audio to process in milliseconds
|
||||||
-mc N, --max-context N [-1 ] maximum number of text context tokens to store
|
-mc N, --max-context N [-1 ] maximum number of text context tokens to store
|
||||||
-ml N, --max-len N [0 ] maximum segment length in characters
|
-ml N, --max-len N [0 ] maximum segment length in characters
|
||||||
-bo N, --best-of N [5 ] number of best candidates to keep
|
|
||||||
-bs N, --beam-size N [-1 ] beam size for beam search
|
|
||||||
-wt N, --word-thold N [0.01 ] word timestamp probability threshold
|
-wt N, --word-thold N [0.01 ] word timestamp probability threshold
|
||||||
-et N, --entropy-thold N [2.40 ] entropy threshold for decoder fail
|
|
||||||
-lpt N, --logprob-thold N [-1.00 ] log probability threshold for decoder fail
|
|
||||||
-su, --speed-up [false ] speed up audio by x2 (reduced accuracy)
|
-su, --speed-up [false ] speed up audio by x2 (reduced accuracy)
|
||||||
-tr, --translate [false ] translate from source language to english
|
-tr, --translate [false ] translate from source language to english
|
||||||
-di, --diarize [false ] stereo audio diarization
|
|
||||||
-nf, --no-fallback [false ] do not use temperature fallback while decoding
|
|
||||||
-otxt, --output-txt [false ] output result in a text file
|
-otxt, --output-txt [false ] output result in a text file
|
||||||
-ovtt, --output-vtt [false ] output result in a vtt file
|
-ovtt, --output-vtt [false ] output result in a vtt file
|
||||||
-osrt, --output-srt [false ] output result in a srt file
|
-osrt, --output-srt [false ] output result in a srt file
|
||||||
-owts, --output-words [false ] output script for generating karaoke video
|
-owts, --output-words [false ] output script for generating karaoke video
|
||||||
-ocsv, --output-csv [false ] output result in a CSV file
|
|
||||||
-of FNAME, --output-file FNAME [ ] output file path (without file extension)
|
|
||||||
-ps, --print-special [false ] print special tokens
|
-ps, --print-special [false ] print special tokens
|
||||||
-pc, --print-colors [false ] print colors
|
-pc, --print-colors [false ] print colors
|
||||||
-pp, --print-progress [false ] print progress
|
|
||||||
-nt, --no-timestamps [true ] do not print timestamps
|
-nt, --no-timestamps [true ] do not print timestamps
|
||||||
-l LANG, --language LANG [en ] spoken language ('auto' for auto-detect)
|
-l LANG, --language LANG [en ] spoken language
|
||||||
--prompt PROMPT [ ] initial prompt
|
|
||||||
-m FNAME, --model FNAME [models/ggml-base.en.bin] model path
|
-m FNAME, --model FNAME [models/ggml-base.en.bin] model path
|
||||||
-f FNAME, --file FNAME [ ] input WAV file path
|
-f FNAME, --file FNAME [ ] input WAV file path
|
||||||
```
|
```
|
||||||
|
|||||||
@ -63,14 +63,12 @@ struct whisper_params {
|
|||||||
int32_t beam_size = -1;
|
int32_t beam_size = -1;
|
||||||
|
|
||||||
float word_thold = 0.01f;
|
float word_thold = 0.01f;
|
||||||
float entropy_thold = 2.40f;
|
float entropy_thold = 2.4f;
|
||||||
float logprob_thold = -1.00f;
|
float logprob_thold = -1.0f;
|
||||||
|
|
||||||
bool speed_up = false;
|
bool speed_up = false;
|
||||||
bool translate = false;
|
bool translate = false;
|
||||||
bool diarize = false;
|
bool diarize = false;
|
||||||
bool split_on_word = false;
|
|
||||||
bool no_fallback = false;
|
|
||||||
bool output_txt = false;
|
bool output_txt = false;
|
||||||
bool output_vtt = false;
|
bool output_vtt = false;
|
||||||
bool output_srt = false;
|
bool output_srt = false;
|
||||||
@ -86,7 +84,6 @@ struct whisper_params {
|
|||||||
std::string model = "models/ggml-base.en.bin";
|
std::string model = "models/ggml-base.en.bin";
|
||||||
|
|
||||||
std::vector<std::string> fname_inp = {};
|
std::vector<std::string> fname_inp = {};
|
||||||
std::vector<std::string> fname_outp = {};
|
|
||||||
};
|
};
|
||||||
|
|
||||||
void whisper_print_usage(int argc, char ** argv, const whisper_params & params);
|
void whisper_print_usage(int argc, char ** argv, const whisper_params & params);
|
||||||
@ -119,14 +116,11 @@ bool whisper_params_parse(int argc, char ** argv, whisper_params & params) {
|
|||||||
else if (arg == "-su" || arg == "--speed-up") { params.speed_up = true; }
|
else if (arg == "-su" || arg == "--speed-up") { params.speed_up = true; }
|
||||||
else if (arg == "-tr" || arg == "--translate") { params.translate = true; }
|
else if (arg == "-tr" || arg == "--translate") { params.translate = true; }
|
||||||
else if (arg == "-di" || arg == "--diarize") { params.diarize = true; }
|
else if (arg == "-di" || arg == "--diarize") { params.diarize = true; }
|
||||||
else if (arg == "-sow" || arg == "--split-on-word") { params.split_on_word = true; }
|
|
||||||
else if (arg == "-nf" || arg == "--no-fallback") { params.no_fallback = true; }
|
|
||||||
else if (arg == "-otxt" || arg == "--output-txt") { params.output_txt = true; }
|
else if (arg == "-otxt" || arg == "--output-txt") { params.output_txt = true; }
|
||||||
else if (arg == "-ovtt" || arg == "--output-vtt") { params.output_vtt = true; }
|
else if (arg == "-ovtt" || arg == "--output-vtt") { params.output_vtt = true; }
|
||||||
else if (arg == "-osrt" || arg == "--output-srt") { params.output_srt = true; }
|
else if (arg == "-osrt" || arg == "--output-srt") { params.output_srt = true; }
|
||||||
else if (arg == "-owts" || arg == "--output-words") { params.output_wts = true; }
|
else if (arg == "-owts" || arg == "--output-words") { params.output_wts = true; }
|
||||||
else if (arg == "-ocsv" || arg == "--output-csv") { params.output_csv = true; }
|
else if (arg == "-ocsv" || arg == "--output-csv") { params.output_csv = true; }
|
||||||
else if (arg == "-of" || arg == "--output-file") { params.fname_outp.emplace_back(argv[++i]); }
|
|
||||||
else if (arg == "-ps" || arg == "--print-special") { params.print_special = true; }
|
else if (arg == "-ps" || arg == "--print-special") { params.print_special = true; }
|
||||||
else if (arg == "-pc" || arg == "--print-colors") { params.print_colors = true; }
|
else if (arg == "-pc" || arg == "--print-colors") { params.print_colors = true; }
|
||||||
else if (arg == "-pp" || arg == "--print-progress") { params.print_progress = true; }
|
else if (arg == "-pp" || arg == "--print-progress") { params.print_progress = true; }
|
||||||
@ -158,7 +152,6 @@ void whisper_print_usage(int /*argc*/, char ** argv, const whisper_params & para
|
|||||||
fprintf(stderr, " -d N, --duration N [%-7d] duration of audio to process in milliseconds\n", params.duration_ms);
|
fprintf(stderr, " -d N, --duration N [%-7d] duration of audio to process in milliseconds\n", params.duration_ms);
|
||||||
fprintf(stderr, " -mc N, --max-context N [%-7d] maximum number of text context tokens to store\n", params.max_context);
|
fprintf(stderr, " -mc N, --max-context N [%-7d] maximum number of text context tokens to store\n", params.max_context);
|
||||||
fprintf(stderr, " -ml N, --max-len N [%-7d] maximum segment length in characters\n", params.max_len);
|
fprintf(stderr, " -ml N, --max-len N [%-7d] maximum segment length in characters\n", params.max_len);
|
||||||
fprintf(stderr, " -sow, --split-on-word [%-7s] split on word rather than on token\n", params.split_on_word ? "true" : "false");
|
|
||||||
fprintf(stderr, " -bo N, --best-of N [%-7d] number of best candidates to keep\n", params.best_of);
|
fprintf(stderr, " -bo N, --best-of N [%-7d] number of best candidates to keep\n", params.best_of);
|
||||||
fprintf(stderr, " -bs N, --beam-size N [%-7d] beam size for beam search\n", params.beam_size);
|
fprintf(stderr, " -bs N, --beam-size N [%-7d] beam size for beam search\n", params.beam_size);
|
||||||
fprintf(stderr, " -wt N, --word-thold N [%-7.2f] word timestamp probability threshold\n", params.word_thold);
|
fprintf(stderr, " -wt N, --word-thold N [%-7.2f] word timestamp probability threshold\n", params.word_thold);
|
||||||
@ -167,13 +160,11 @@ void whisper_print_usage(int /*argc*/, char ** argv, const whisper_params & para
|
|||||||
fprintf(stderr, " -su, --speed-up [%-7s] speed up audio by x2 (reduced accuracy)\n", params.speed_up ? "true" : "false");
|
fprintf(stderr, " -su, --speed-up [%-7s] speed up audio by x2 (reduced accuracy)\n", params.speed_up ? "true" : "false");
|
||||||
fprintf(stderr, " -tr, --translate [%-7s] translate from source language to english\n", params.translate ? "true" : "false");
|
fprintf(stderr, " -tr, --translate [%-7s] translate from source language to english\n", params.translate ? "true" : "false");
|
||||||
fprintf(stderr, " -di, --diarize [%-7s] stereo audio diarization\n", params.diarize ? "true" : "false");
|
fprintf(stderr, " -di, --diarize [%-7s] stereo audio diarization\n", params.diarize ? "true" : "false");
|
||||||
fprintf(stderr, " -nf, --no-fallback [%-7s] do not use temperature fallback while decoding\n", params.no_fallback ? "true" : "false");
|
|
||||||
fprintf(stderr, " -otxt, --output-txt [%-7s] output result in a text file\n", params.output_txt ? "true" : "false");
|
fprintf(stderr, " -otxt, --output-txt [%-7s] output result in a text file\n", params.output_txt ? "true" : "false");
|
||||||
fprintf(stderr, " -ovtt, --output-vtt [%-7s] output result in a vtt file\n", params.output_vtt ? "true" : "false");
|
fprintf(stderr, " -ovtt, --output-vtt [%-7s] output result in a vtt file\n", params.output_vtt ? "true" : "false");
|
||||||
fprintf(stderr, " -osrt, --output-srt [%-7s] output result in a srt file\n", params.output_srt ? "true" : "false");
|
fprintf(stderr, " -osrt, --output-srt [%-7s] output result in a srt file\n", params.output_srt ? "true" : "false");
|
||||||
fprintf(stderr, " -owts, --output-words [%-7s] output script for generating karaoke video\n", params.output_wts ? "true" : "false");
|
fprintf(stderr, " -owts, --output-words [%-7s] output script for generating karaoke video\n", params.output_wts ? "true" : "false");
|
||||||
fprintf(stderr, " -ocsv, --output-csv [%-7s] output result in a CSV file\n", params.output_csv ? "true" : "false");
|
fprintf(stderr, " -ocsv, --output-csv [%-7s] output result in a CSV file\n", params.output_csv ? "true" : "false");
|
||||||
fprintf(stderr, " -of FNAME, --output-file FNAME [%-7s] output file path (without file extension)\n", "");
|
|
||||||
fprintf(stderr, " -ps, --print-special [%-7s] print special tokens\n", params.print_special ? "true" : "false");
|
fprintf(stderr, " -ps, --print-special [%-7s] print special tokens\n", params.print_special ? "true" : "false");
|
||||||
fprintf(stderr, " -pc, --print-colors [%-7s] print colors\n", params.print_colors ? "true" : "false");
|
fprintf(stderr, " -pc, --print-colors [%-7s] print colors\n", params.print_colors ? "true" : "false");
|
||||||
fprintf(stderr, " -pp, --print-progress [%-7s] print progress\n", params.print_progress ? "true" : "false");
|
fprintf(stderr, " -pp, --print-progress [%-7s] print progress\n", params.print_progress ? "true" : "false");
|
||||||
@ -352,6 +343,9 @@ bool output_csv(struct whisper_context * ctx, const char * fname) {
|
|||||||
const int n_segments = whisper_full_n_segments(ctx);
|
const int n_segments = whisper_full_n_segments(ctx);
|
||||||
for (int i = 0; i < n_segments; ++i) {
|
for (int i = 0; i < n_segments; ++i) {
|
||||||
const char * text = whisper_full_get_segment_text(ctx, i);
|
const char * text = whisper_full_get_segment_text(ctx, i);
|
||||||
|
if (text[0] == ' ') {
|
||||||
|
text = text + sizeof(char); //whisper_full_get_segment_text() returns a string with leading space, point to the next character.
|
||||||
|
}
|
||||||
const int64_t t0 = whisper_full_get_segment_t0(ctx, i);
|
const int64_t t0 = whisper_full_get_segment_t0(ctx, i);
|
||||||
const int64_t t1 = whisper_full_get_segment_t1(ctx, i);
|
const int64_t t1 = whisper_full_get_segment_t1(ctx, i);
|
||||||
|
|
||||||
@ -520,7 +514,6 @@ int main(int argc, char ** argv) {
|
|||||||
|
|
||||||
for (int f = 0; f < (int) params.fname_inp.size(); ++f) {
|
for (int f = 0; f < (int) params.fname_inp.size(); ++f) {
|
||||||
const auto fname_inp = params.fname_inp[f];
|
const auto fname_inp = params.fname_inp[f];
|
||||||
const auto fname_outp = f < (int) params.fname_outp.size() && !params.fname_outp[f].empty() ? params.fname_outp[f] : params.fname_inp[f];
|
|
||||||
|
|
||||||
std::vector<float> pcmf32; // mono-channel F32 PCM
|
std::vector<float> pcmf32; // mono-channel F32 PCM
|
||||||
std::vector<std::vector<float>> pcmf32s; // stereo-channel F32 PCM
|
std::vector<std::vector<float>> pcmf32s; // stereo-channel F32 PCM
|
||||||
@ -653,20 +646,18 @@ int main(int argc, char ** argv) {
|
|||||||
|
|
||||||
wparams.token_timestamps = params.output_wts || params.max_len > 0;
|
wparams.token_timestamps = params.output_wts || params.max_len > 0;
|
||||||
wparams.thold_pt = params.word_thold;
|
wparams.thold_pt = params.word_thold;
|
||||||
|
wparams.entropy_thold = params.entropy_thold;
|
||||||
|
wparams.logprob_thold = params.logprob_thold;
|
||||||
wparams.max_len = params.output_wts && params.max_len == 0 ? 60 : params.max_len;
|
wparams.max_len = params.output_wts && params.max_len == 0 ? 60 : params.max_len;
|
||||||
wparams.split_on_word = params.split_on_word;
|
|
||||||
|
|
||||||
wparams.speed_up = params.speed_up;
|
wparams.speed_up = params.speed_up;
|
||||||
|
|
||||||
wparams.prompt_tokens = prompt_tokens.empty() ? nullptr : prompt_tokens.data();
|
|
||||||
wparams.prompt_n_tokens = prompt_tokens.empty() ? 0 : prompt_tokens.size();
|
|
||||||
|
|
||||||
wparams.greedy.best_of = params.best_of;
|
wparams.greedy.best_of = params.best_of;
|
||||||
wparams.beam_search.beam_size = params.beam_size;
|
wparams.beam_search.beam_size = params.beam_size;
|
||||||
|
wparams.temperature_inc = -1;
|
||||||
|
|
||||||
wparams.temperature_inc = params.no_fallback ? 0.0f : wparams.temperature_inc;
|
wparams.prompt_tokens = prompt_tokens.empty() ? nullptr : prompt_tokens.data();
|
||||||
wparams.entropy_thold = params.entropy_thold;
|
wparams.prompt_n_tokens = prompt_tokens.empty() ? 0 : prompt_tokens.size();
|
||||||
wparams.logprob_thold = params.logprob_thold;
|
|
||||||
|
|
||||||
whisper_print_user_data user_data = { ¶ms, &pcmf32s };
|
whisper_print_user_data user_data = { ¶ms, &pcmf32s };
|
||||||
|
|
||||||
@ -701,31 +692,31 @@ int main(int argc, char ** argv) {
|
|||||||
|
|
||||||
// output to text file
|
// output to text file
|
||||||
if (params.output_txt) {
|
if (params.output_txt) {
|
||||||
const auto fname_txt = fname_outp + ".txt";
|
const auto fname_txt = fname_inp + ".txt";
|
||||||
output_txt(ctx, fname_txt.c_str());
|
output_txt(ctx, fname_txt.c_str());
|
||||||
}
|
}
|
||||||
|
|
||||||
// output to VTT file
|
// output to VTT file
|
||||||
if (params.output_vtt) {
|
if (params.output_vtt) {
|
||||||
const auto fname_vtt = fname_outp + ".vtt";
|
const auto fname_vtt = fname_inp + ".vtt";
|
||||||
output_vtt(ctx, fname_vtt.c_str());
|
output_vtt(ctx, fname_vtt.c_str());
|
||||||
}
|
}
|
||||||
|
|
||||||
// output to SRT file
|
// output to SRT file
|
||||||
if (params.output_srt) {
|
if (params.output_srt) {
|
||||||
const auto fname_srt = fname_outp + ".srt";
|
const auto fname_srt = fname_inp + ".srt";
|
||||||
output_srt(ctx, fname_srt.c_str(), params);
|
output_srt(ctx, fname_srt.c_str(), params);
|
||||||
}
|
}
|
||||||
|
|
||||||
// output to WTS file
|
// output to WTS file
|
||||||
if (params.output_wts) {
|
if (params.output_wts) {
|
||||||
const auto fname_wts = fname_outp + ".wts";
|
const auto fname_wts = fname_inp + ".wts";
|
||||||
output_wts(ctx, fname_wts.c_str(), fname_inp.c_str(), params, float(pcmf32.size() + 1000)/WHISPER_SAMPLE_RATE);
|
output_wts(ctx, fname_wts.c_str(), fname_inp.c_str(), params, float(pcmf32.size() + 1000)/WHISPER_SAMPLE_RATE);
|
||||||
}
|
}
|
||||||
|
|
||||||
// output to CSV file
|
// output to CSV file
|
||||||
if (params.output_csv) {
|
if (params.output_csv) {
|
||||||
const auto fname_csv = fname_outp + ".csv";
|
const auto fname_csv = fname_inp + ".csv";
|
||||||
output_csv(ctx, fname_csv.c_str());
|
output_csv(ctx, fname_csv.c_str());
|
||||||
}
|
}
|
||||||
|
|
||||||
|
|||||||
@ -423,8 +423,7 @@ int main(int argc, char ** argv) {
|
|||||||
return 1;
|
return 1;
|
||||||
}
|
}
|
||||||
|
|
||||||
params.keep_ms = std::min(params.keep_ms, params.step_ms);
|
params.keep_ms = std::min(params.keep_ms, params.step_ms); // cannot be more than step_ms
|
||||||
params.length_ms = std::max(params.length_ms, params.step_ms);
|
|
||||||
|
|
||||||
const int n_samples_step = (params.step_ms *1e-3)*WHISPER_SAMPLE_RATE;
|
const int n_samples_step = (params.step_ms *1e-3)*WHISPER_SAMPLE_RATE;
|
||||||
const int n_samples_len = (params.length_ms*1e-3)*WHISPER_SAMPLE_RATE;
|
const int n_samples_len = (params.length_ms*1e-3)*WHISPER_SAMPLE_RATE;
|
||||||
@ -433,7 +432,7 @@ int main(int argc, char ** argv) {
|
|||||||
|
|
||||||
const bool use_vad = n_samples_step <= 0; // sliding window mode uses VAD
|
const bool use_vad = n_samples_step <= 0; // sliding window mode uses VAD
|
||||||
|
|
||||||
const int n_new_line = !use_vad ? std::max(1, params.length_ms / params.step_ms - 1) : 1; // number of steps to print new line
|
const int n_new_line = !use_vad ? params.length_ms / params.step_ms - 1 : 1; // number of steps to print new line
|
||||||
|
|
||||||
params.no_timestamps = !use_vad;
|
params.no_timestamps = !use_vad;
|
||||||
params.no_context |= use_vad;
|
params.no_context |= use_vad;
|
||||||
|
|||||||
@ -32,8 +32,8 @@ set_target_properties(${TARGET} PROPERTIES LINK_FLAGS " \
|
|||||||
--bind \
|
--bind \
|
||||||
-s USE_PTHREADS=1 \
|
-s USE_PTHREADS=1 \
|
||||||
-s PTHREAD_POOL_SIZE=8 \
|
-s PTHREAD_POOL_SIZE=8 \
|
||||||
-s INITIAL_MEMORY=1500MB \
|
-s INITIAL_MEMORY=1024MB \
|
||||||
-s TOTAL_MEMORY=1500MB \
|
-s TOTAL_MEMORY=1024MB \
|
||||||
-s FORCE_FILESYSTEM=1 \
|
-s FORCE_FILESYSTEM=1 \
|
||||||
-s EXPORTED_RUNTIME_METHODS=\"['print', 'printErr', 'ccall', 'cwrap']\" \
|
-s EXPORTED_RUNTIME_METHODS=\"['print', 'printErr', 'ccall', 'cwrap']\" \
|
||||||
${EXTRA_FLAGS} \
|
${EXTRA_FLAGS} \
|
||||||
|
|||||||
@ -50,8 +50,6 @@
|
|||||||
<button id="fetch-whisper-tiny" onclick="loadWhisper('tiny')">tiny (75 MB)</button>
|
<button id="fetch-whisper-tiny" onclick="loadWhisper('tiny')">tiny (75 MB)</button>
|
||||||
<button id="fetch-whisper-base-en" onclick="loadWhisper('base.en')">base.en (142 MB)</button>
|
<button id="fetch-whisper-base-en" onclick="loadWhisper('base.en')">base.en (142 MB)</button>
|
||||||
<button id="fetch-whisper-base" onclick="loadWhisper('base')">base (142 MB)</button>
|
<button id="fetch-whisper-base" onclick="loadWhisper('base')">base (142 MB)</button>
|
||||||
<button id="fetch-whisper-small-en" onclick="loadWhisper('small.en')">small.en (466 MB)</button>
|
|
||||||
<button id="fetch-whisper-small" onclick="loadWhisper('small')">small (466 MB)</button>
|
|
||||||
<span id="fetch-whisper-progress"></span>
|
<span id="fetch-whisper-progress"></span>
|
||||||
|
|
||||||
<input type="file" id="whisper-file" name="file" onchange="loadFile(event, 'whisper.bin')" />
|
<input type="file" id="whisper-file" name="file" onchange="loadFile(event, 'whisper.bin')" />
|
||||||
@ -62,8 +60,8 @@
|
|||||||
<!-- radio button to select between file upload or microphone -->
|
<!-- radio button to select between file upload or microphone -->
|
||||||
<div id="input">
|
<div id="input">
|
||||||
Input:
|
Input:
|
||||||
<input type="radio" id="file" name="input" value="file" checked="checked" onchange="changeInput('file')" /> <label for="file">File</label>
|
<input type="radio" id="file" name="input" value="file" checked="checked" onchange="changeInput('file')" /> File
|
||||||
<input type="radio" id="mic" name="input" value="mic" onchange="changeInput('mic')" /> <label for="mic">Microphone</label>
|
<input type="radio" id="mic" name="input" value="mic" onchange="changeInput('mic')" /> Microphone
|
||||||
</div>
|
</div>
|
||||||
|
|
||||||
<br>
|
<br>
|
||||||
@ -288,10 +286,8 @@
|
|||||||
|
|
||||||
document.getElementById('fetch-whisper-tiny-en').style.display = 'none';
|
document.getElementById('fetch-whisper-tiny-en').style.display = 'none';
|
||||||
document.getElementById('fetch-whisper-base-en').style.display = 'none';
|
document.getElementById('fetch-whisper-base-en').style.display = 'none';
|
||||||
document.getElementById('fetch-whisper-small-en').style.display = 'none';
|
|
||||||
document.getElementById('fetch-whisper-tiny' ).style.display = 'none';
|
document.getElementById('fetch-whisper-tiny' ).style.display = 'none';
|
||||||
document.getElementById('fetch-whisper-base' ).style.display = 'none';
|
document.getElementById('fetch-whisper-base' ).style.display = 'none';
|
||||||
document.getElementById('fetch-whisper-small' ).style.display = 'none';
|
|
||||||
document.getElementById('whisper-file' ).style.display = 'none';
|
document.getElementById('whisper-file' ).style.display = 'none';
|
||||||
document.getElementById('model-whisper-status' ).innerHTML = 'loaded model: ' + file.name;
|
document.getElementById('model-whisper-status' ).innerHTML = 'loaded model: ' + file.name;
|
||||||
}
|
}
|
||||||
@ -302,8 +298,6 @@
|
|||||||
'tiny': 'https://whisper.ggerganov.com/ggml-model-whisper-tiny.bin',
|
'tiny': 'https://whisper.ggerganov.com/ggml-model-whisper-tiny.bin',
|
||||||
'base.en': 'https://whisper.ggerganov.com/ggml-model-whisper-base.en.bin',
|
'base.en': 'https://whisper.ggerganov.com/ggml-model-whisper-base.en.bin',
|
||||||
'base': 'https://whisper.ggerganov.com/ggml-model-whisper-base.bin',
|
'base': 'https://whisper.ggerganov.com/ggml-model-whisper-base.bin',
|
||||||
'small.en': 'https://whisper.ggerganov.com/ggml-model-whisper-small.en.bin',
|
|
||||||
'small': 'https://whisper.ggerganov.com/ggml-model-whisper-small.bin',
|
|
||||||
};
|
};
|
||||||
|
|
||||||
let sizes = {
|
let sizes = {
|
||||||
@ -311,8 +305,6 @@
|
|||||||
'tiny': 75,
|
'tiny': 75,
|
||||||
'base.en': 142,
|
'base.en': 142,
|
||||||
'base': 142,
|
'base': 142,
|
||||||
'small.en': 466,
|
|
||||||
'small': 466,
|
|
||||||
};
|
};
|
||||||
|
|
||||||
let url = urls[model];
|
let url = urls[model];
|
||||||
@ -323,10 +315,8 @@
|
|||||||
|
|
||||||
document.getElementById('fetch-whisper-tiny-en').style.display = 'none';
|
document.getElementById('fetch-whisper-tiny-en').style.display = 'none';
|
||||||
document.getElementById('fetch-whisper-base-en').style.display = 'none';
|
document.getElementById('fetch-whisper-base-en').style.display = 'none';
|
||||||
document.getElementById('fetch-whisper-small-en').style.display = 'none';
|
|
||||||
document.getElementById('fetch-whisper-tiny' ).style.display = 'none';
|
document.getElementById('fetch-whisper-tiny' ).style.display = 'none';
|
||||||
document.getElementById('fetch-whisper-base' ).style.display = 'none';
|
document.getElementById('fetch-whisper-base' ).style.display = 'none';
|
||||||
document.getElementById('fetch-whisper-small' ).style.display = 'none';
|
|
||||||
document.getElementById('whisper-file' ).style.display = 'none';
|
document.getElementById('whisper-file' ).style.display = 'none';
|
||||||
document.getElementById('model-whisper-status' ).innerHTML = 'loading model: ' + model;
|
document.getElementById('model-whisper-status' ).innerHTML = 'loading model: ' + model;
|
||||||
|
|
||||||
@ -339,10 +329,8 @@
|
|||||||
var el;
|
var el;
|
||||||
el = document.getElementById('fetch-whisper-tiny-en'); if (el) el.style.display = 'inline-block';
|
el = document.getElementById('fetch-whisper-tiny-en'); if (el) el.style.display = 'inline-block';
|
||||||
el = document.getElementById('fetch-whisper-base-en'); if (el) el.style.display = 'inline-block';
|
el = document.getElementById('fetch-whisper-base-en'); if (el) el.style.display = 'inline-block';
|
||||||
el = document.getElementById('fetch-whisper-small-en'); if (el) el.style.display = 'inline-block';
|
|
||||||
el = document.getElementById('fetch-whisper-tiny' ); if (el) el.style.display = 'inline-block';
|
el = document.getElementById('fetch-whisper-tiny' ); if (el) el.style.display = 'inline-block';
|
||||||
el = document.getElementById('fetch-whisper-base' ); if (el) el.style.display = 'inline-block';
|
el = document.getElementById('fetch-whisper-base' ); if (el) el.style.display = 'inline-block';
|
||||||
el = document.getElementById('fetch-whisper-small' ); if (el) el.style.display = 'inline-block';
|
|
||||||
el = document.getElementById('whisper-file' ); if (el) el.style.display = 'inline-block';
|
el = document.getElementById('whisper-file' ); if (el) el.style.display = 'inline-block';
|
||||||
el = document.getElementById('model-whisper-status' ); if (el) el.innerHTML = '';
|
el = document.getElementById('model-whisper-status' ); if (el) el.innerHTML = '';
|
||||||
};
|
};
|
||||||
|
|||||||
@ -1,10 +1,20 @@
|
|||||||
#!/usr/bin/env bash
|
#!/usr/bin/env bash
|
||||||
# shellcheck disable=2086
|
|
||||||
|
# Small shell script to more easily automatically download and transcribe live stream VODs.
|
||||||
|
# This uses YT-DLP, ffmpeg and the CPP version of Whisper: https://github.com/ggerganov/whisper.cpp
|
||||||
|
# Use `./examples/yt-wsp.sh help` to print help info.
|
||||||
|
#
|
||||||
|
# Sample usage:
|
||||||
|
#
|
||||||
|
# git clone https://github.com/ggerganov/whisper.cpp
|
||||||
|
# cd whisper.cpp
|
||||||
|
# make
|
||||||
|
# ./examples/yt-wsp.sh https://www.youtube.com/watch?v=1234567890
|
||||||
|
#
|
||||||
|
|
||||||
# MIT License
|
# MIT License
|
||||||
|
|
||||||
# Copyright (c) 2022 Daniils Petrovs
|
# Copyright (c) 2022 Daniils Petrovs
|
||||||
# Copyright (c) 2023 Jennifer Capasso
|
|
||||||
|
|
||||||
# Permission is hereby granted, free of charge, to any person obtaining a copy
|
# Permission is hereby granted, free of charge, to any person obtaining a copy
|
||||||
# of this software and associated documentation files (the "Software"), to deal
|
# of this software and associated documentation files (the "Software"), to deal
|
||||||
@ -24,181 +34,114 @@
|
|||||||
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
||||||
# SOFTWARE.
|
# SOFTWARE.
|
||||||
|
|
||||||
# Small shell script to more easily automatically download and transcribe live stream VODs.
|
|
||||||
# This uses YT-DLP, ffmpeg and the CPP version of Whisper: https://github.com/ggerganov/whisper.cpp
|
|
||||||
# Use `./examples/yt-wsp.sh help` to print help info.
|
|
||||||
#
|
|
||||||
# Sample usage:
|
|
||||||
#
|
|
||||||
# git clone https://github.com/ggerganov/whisper.cpp
|
|
||||||
# cd whisper.cpp
|
|
||||||
# make
|
|
||||||
# ./examples/yt-wsp.sh https://www.youtube.com/watch?v=1234567890
|
|
||||||
#
|
|
||||||
|
|
||||||
set -Eeuo pipefail
|
set -Eeuo pipefail
|
||||||
|
|
||||||
# get script file location
|
# You can find how to download models in the OG repo: https://github.com/ggerganov/whisper.cpp/#usage
|
||||||
SCRIPT_PATH="$(realpath -e ${BASH_SOURCE[0]})";
|
MODEL_PATH="${MODEL_PATH:-models/ggml-base.en.bin}" # Set to a multilingual model if you want to translate from foreign lang to en
|
||||||
SCRIPT_DIR="${SCRIPT_PATH%/*}"
|
WHISPER_EXECUTABLE="${WHISPER_EXECUTABLE:-whisper}" # Where to find the whisper.cpp executable
|
||||||
|
WHISPER_LANG="${WHISPER_LANG:-en}" # Set to desired lang to translate from
|
||||||
################################################################################
|
|
||||||
# Documentation on downloading models can be found in the whisper.cpp repo:
|
|
||||||
# https://github.com/ggerganov/whisper.cpp/#usage
|
|
||||||
#
|
|
||||||
# note: unless a multilingual model is specified, WHISPER_LANG will be ignored
|
|
||||||
# and the video will be transcribed as if the audio were in the English language
|
|
||||||
################################################################################
|
|
||||||
MODEL_PATH="${MODEL_PATH:-${SCRIPT_DIR}/../models/ggml-base.en.bin}"
|
|
||||||
|
|
||||||
################################################################################
|
|
||||||
# Where to find the whisper.cpp executable. default to the examples directory
|
|
||||||
# which holds this script in source control
|
|
||||||
################################################################################
|
|
||||||
WHISPER_EXECUTABLE="${WHISPER_EXECUTABLE:-${SCRIPT_DIR}/../main}";
|
|
||||||
|
|
||||||
# Set to desired language to be translated into english
|
|
||||||
WHISPER_LANG="${WHISPER_LANG:-en}";
|
|
||||||
|
|
||||||
# Default to 4 threads (this was most performant on my 2020 M1 MBP)
|
|
||||||
WHISPER_THREAD_COUNT="${WHISPER_THREAD_COUNT:-4}";
|
|
||||||
|
|
||||||
msg() {
|
msg() {
|
||||||
echo >&2 -e "${1-}"
|
echo >&2 -e "${1-}"
|
||||||
}
|
}
|
||||||
|
|
||||||
################################################################################
|
|
||||||
# create a temporary directory to work in
|
|
||||||
# set the temp_dir and temp_filename variables
|
|
||||||
################################################################################
|
|
||||||
temp_dir="$(mktemp -d ${SCRIPT_DIR}/tmp.XXXXXX)";
|
|
||||||
temp_filename="${temp_dir}/yt-dlp-filename";
|
|
||||||
|
|
||||||
################################################################################
|
|
||||||
# for now we only take one argument
|
|
||||||
# TODO: a for loop
|
|
||||||
################################################################################
|
|
||||||
source_url="${1}"
|
|
||||||
|
|
||||||
|
|
||||||
title_name="";
|
|
||||||
|
|
||||||
|
|
||||||
cleanup() {
|
cleanup() {
|
||||||
local -r clean_me="${1}";
|
msg "Cleaning up..."
|
||||||
|
rm -rf "${temp_dir}" "vod-resampled.wav" "vod-resampled.wav.srt"
|
||||||
if [ -d "${clean_me}" ]; then
|
|
||||||
msg "Cleaning up...";
|
|
||||||
rm -rf "${clean_me}";
|
|
||||||
else
|
|
||||||
msg "'${clean_me}' does not appear to be a directory!";
|
|
||||||
exit 1;
|
|
||||||
fi;
|
|
||||||
}
|
}
|
||||||
|
|
||||||
print_help() {
|
print_help() {
|
||||||
echo "################################################################################"
|
|
||||||
echo "Usage: ./examples/yt-wsp.sh <video_url>"
|
echo "Usage: ./examples/yt-wsp.sh <video_url>"
|
||||||
echo "# See configurable env variables in the script; there are many!"
|
echo "See configurable env variables in the script"
|
||||||
echo "# This script will produce an MP4 muxed file in the working directory; it will"
|
echo "This will produce an MP4 muxed file called res.mp4 in the working directory"
|
||||||
echo "# be named for the title and id of the video."
|
echo "Requirements: ffmpeg yt-dlp whisper"
|
||||||
echo "# passing in https://youtu.be/VYJtb2YXae8 produces a file named";
|
echo "Whisper needs to be built into the main binary with make, then you can rename it to something like 'whisper' and add it to your PATH for convenience."
|
||||||
echo "# 'Why_we_all_need_subtitles_now-VYJtb2YXae8-res.mp4'"
|
echo "E.g. in the root of Whisper.cpp, run: 'make && cp ./main /usr/local/bin/whisper'"
|
||||||
echo "# Requirements: ffmpeg yt-dlp whisper.cpp"
|
|
||||||
echo "################################################################################"
|
|
||||||
}
|
}
|
||||||
|
|
||||||
check_requirements() {
|
check_requirements() {
|
||||||
if ! command -v ffmpeg &>/dev/null; then
|
if ! command -v ffmpeg &>/dev/null; then
|
||||||
echo "ffmpeg is required: https://ffmpeg.org";
|
echo "ffmpeg is required (https://ffmpeg.org)."
|
||||||
exit 1
|
exit 1
|
||||||
fi;
|
fi
|
||||||
|
|
||||||
if ! command -v yt-dlp &>/dev/null; then
|
if ! command -v yt-dlp &>/dev/null; then
|
||||||
echo "yt-dlp is required: https://github.com/yt-dlp/yt-dlp";
|
echo "yt-dlp is required (https://github.com/yt-dlp/yt-dlp)."
|
||||||
exit 1;
|
exit 1
|
||||||
fi;
|
fi
|
||||||
|
|
||||||
if ! command -v "${WHISPER_EXECUTABLE}" &>/dev/null; then
|
|
||||||
echo "The C++ implementation of Whisper is required: https://github.com/ggerganov/whisper.cpp"
|
|
||||||
echo "Sample usage:";
|
|
||||||
echo "";
|
|
||||||
echo " git clone https://github.com/ggerganov/whisper.cpp";
|
|
||||||
echo " cd whisper.cpp";
|
|
||||||
echo " make";
|
|
||||||
echo " ./examples/yt-wsp.sh https://www.youtube.com/watch?v=1234567890";
|
|
||||||
echo "";
|
|
||||||
exit 1;
|
|
||||||
fi;
|
|
||||||
|
|
||||||
|
if ! command -v "$WHISPER_EXECUTABLE" &>/dev/null; then
|
||||||
|
WHISPER_EXECUTABLE="./main"
|
||||||
|
if ! command -v "$WHISPER_EXECUTABLE" &>/dev/null; then
|
||||||
|
echo "Whisper is required (https://github.com/ggerganov/whisper.cpp):"
|
||||||
|
echo "Sample usage:"
|
||||||
|
echo ""
|
||||||
|
echo " git clone https://github.com/ggerganov/whisper.cpp"
|
||||||
|
echo " cd whisper.cpp"
|
||||||
|
echo " make"
|
||||||
|
echo " ./examples/yt-wsp.sh https://www.youtube.com/watch?v=1234567890"
|
||||||
|
echo ""
|
||||||
|
exit 1
|
||||||
|
fi
|
||||||
|
fi
|
||||||
}
|
}
|
||||||
|
|
||||||
if [[ "${#}" -lt 1 ]]; then
|
if [[ $# -lt 1 ]]; then
|
||||||
print_help;
|
print_help
|
||||||
exit 1;
|
exit 1
|
||||||
fi
|
fi
|
||||||
|
|
||||||
if [[ "${1##-*}" == "help" ]]; then
|
if [[ "$1" == "help" ]]; then
|
||||||
print_help;
|
print_help
|
||||||
exit 0;
|
exit 0
|
||||||
fi
|
fi
|
||||||
|
|
||||||
check_requirements;
|
temp_dir="tmp"
|
||||||
|
source_url="$1"
|
||||||
|
|
||||||
msg "Downloading VOD...";
|
check_requirements
|
||||||
|
|
||||||
################################################################################
|
msg "Downloading VOD..."
|
||||||
# Download the video, put the dynamic output filename into a variable.
|
|
||||||
# Optionally add --cookies-from-browser BROWSER[+KEYRING][:PROFILE][::CONTAINER]
|
# Optionally add --cookies-from-browser BROWSER[+KEYRING][:PROFILE][::CONTAINER] for members only VODs
|
||||||
# for videos only available to logged-in users.
|
|
||||||
################################################################################
|
|
||||||
yt-dlp \
|
yt-dlp \
|
||||||
-f "bestvideo[ext=mp4]+bestaudio[ext=m4a]/best[ext=mp4]/best" \
|
-f "bestvideo[ext=mp4]+bestaudio[ext=m4a]/best[ext=mp4]/best" \
|
||||||
-o "${temp_dir}/%(title)s-%(id)s.vod.mp4" \
|
|
||||||
--print-to-file "%(filename)s" "${temp_filename}" \
|
|
||||||
--no-simulate \
|
|
||||||
--no-write-auto-subs \
|
|
||||||
--restrict-filenames \
|
|
||||||
--embed-thumbnail \
|
--embed-thumbnail \
|
||||||
--embed-chapters \
|
--embed-chapters \
|
||||||
--xattrs \
|
--xattrs \
|
||||||
"${source_url}";
|
"${source_url}" -o "${temp_dir}/vod.mp4"
|
||||||
|
|
||||||
title_name="$(xargs basename -s .vod.mp4 < ${temp_filename})";
|
msg "Extracting audio and resampling..."
|
||||||
|
|
||||||
msg "Extracting audio and resampling...";
|
ffmpeg -i "${temp_dir}/vod.mp4" \
|
||||||
|
|
||||||
ffmpeg -i "${temp_dir}/${title_name}.vod.mp4" \
|
|
||||||
-hide_banner \
|
-hide_banner \
|
||||||
-vn \
|
|
||||||
-loglevel error \
|
-loglevel error \
|
||||||
-ar 16000 \
|
-ar 16000 \
|
||||||
-ac 1 \
|
-ac 1 \
|
||||||
-c:a pcm_s16le \
|
-c:a \
|
||||||
-y \
|
pcm_s16le -y "vod-resampled.wav"
|
||||||
"${temp_dir}/${title_name}.vod-resampled.wav";
|
|
||||||
|
|
||||||
msg "Transcribing to subtitle file...";
|
msg "Transcribing to subtitle file..."
|
||||||
msg "Whisper specified at: '${WHISPER_EXECUTABLE}'";
|
msg "Whisper specified at: ${WHISPER_EXECUTABLE}"
|
||||||
|
|
||||||
"${WHISPER_EXECUTABLE}" \
|
$WHISPER_EXECUTABLE \
|
||||||
-m "${MODEL_PATH}" \
|
-m "${MODEL_PATH}" \
|
||||||
-l "${WHISPER_LANG}" \
|
-l "${WHISPER_LANG}" \
|
||||||
-f "${temp_dir}/${title_name}.vod-resampled.wav" \
|
-f "vod-resampled.wav" \
|
||||||
-t "${WHISPER_THREAD_COUNT}" \
|
-t 8 \
|
||||||
-osrt \
|
-osrt \
|
||||||
--translate;
|
--translate
|
||||||
|
|
||||||
msg "Embedding subtitle track...";
|
msg "Embedding subtitle track..."
|
||||||
|
|
||||||
ffmpeg -i "${temp_dir}/${title_name}.vod.mp4" \
|
ffmpeg -i "${temp_dir}/vod.mp4" \
|
||||||
-hide_banner \
|
-hide_banner \
|
||||||
-loglevel error \
|
-loglevel error \
|
||||||
-i "${temp_dir}/${title_name}.vod-resampled.wav.srt" \
|
-i "vod-resampled.wav.srt" \
|
||||||
-c copy \
|
-c copy \
|
||||||
-c:s mov_text \
|
-c:s mov_text \
|
||||||
-y "${title_name}-res.mp4";
|
-y res.mp4
|
||||||
|
|
||||||
cleanup "${temp_dir}";
|
cleanup
|
||||||
|
|
||||||
msg "Done! Your finished file is ready: ${title_name}-res.mp4";
|
msg "Done! Your finished file is ready: res.mp4"
|
||||||
|
|||||||
@ -19,7 +19,7 @@ printf "\n"
|
|||||||
./bench -w 1 -t 1 2>&1
|
./bench -w 1 -t 1 2>&1
|
||||||
|
|
||||||
printf "\n"
|
printf "\n"
|
||||||
printf "Running ggml_mul_mat benchmark with $n_threads threads\n"
|
printf "Running ggml_mul_mat benchmark with " $n_threads " threads\n"
|
||||||
printf "\n"
|
printf "\n"
|
||||||
|
|
||||||
./bench -w 2 -t $n_threads 2>&1
|
./bench -w 2 -t $n_threads 2>&1
|
||||||
|
|||||||
114
ggml.c
114
ggml.c
@ -339,12 +339,8 @@ int64_t ggml_cycles_per_ms(void) {
|
|||||||
#if defined(__cpp_lib_hardware_interference_size)
|
#if defined(__cpp_lib_hardware_interference_size)
|
||||||
#define CACHE_LINE_SIZE hardware_destructive_interference_size
|
#define CACHE_LINE_SIZE hardware_destructive_interference_size
|
||||||
#else
|
#else
|
||||||
#if defined(__POWER9_VECTOR__)
|
|
||||||
#define CACHE_LINE_SIZE 128
|
|
||||||
#else
|
|
||||||
#define CACHE_LINE_SIZE 64
|
#define CACHE_LINE_SIZE 64
|
||||||
#endif
|
#endif
|
||||||
#endif
|
|
||||||
|
|
||||||
static const size_t CACHE_LINE_SIZE_F32 = CACHE_LINE_SIZE/sizeof(float);
|
static const size_t CACHE_LINE_SIZE_F32 = CACHE_LINE_SIZE/sizeof(float);
|
||||||
|
|
||||||
@ -613,12 +609,9 @@ static const size_t CACHE_LINE_SIZE_F32 = CACHE_LINE_SIZE/sizeof(float);
|
|||||||
#define GGML_F16_VEC_LOAD(p, i) (i & 0x1) ? \
|
#define GGML_F16_VEC_LOAD(p, i) (i & 0x1) ? \
|
||||||
vec_extract_fp32_from_shorth(vec_xl(0, p - GGML_F16_EPR)) : \
|
vec_extract_fp32_from_shorth(vec_xl(0, p - GGML_F16_EPR)) : \
|
||||||
vec_extract_fp32_from_shortl(vec_xl(0, p))
|
vec_extract_fp32_from_shortl(vec_xl(0, p))
|
||||||
#define GGML_ENDIAN_BYTE(i) ((unsigned char *)&(uint16_t){1})[i]
|
|
||||||
#define GGML_F16_VEC_STORE(p, r, i) \
|
#define GGML_F16_VEC_STORE(p, r, i) \
|
||||||
if (i & 0x1) \
|
if (i & 0x1) \
|
||||||
vec_xst(vec_pack_to_short_fp32(r[i - GGML_ENDIAN_BYTE(1)], \
|
vec_xst(vec_pack_to_short_fp32(r[i], r[i - 1]), 0, p - GGML_F16_EPR)
|
||||||
r[i - GGML_ENDIAN_BYTE(0)]), \
|
|
||||||
0, p - GGML_F16_EPR)
|
|
||||||
|
|
||||||
#elif defined(__wasm_simd128__)
|
#elif defined(__wasm_simd128__)
|
||||||
|
|
||||||
@ -1258,7 +1251,7 @@ static const char * GGML_OP_SYMBOL[GGML_OP_COUNT] = {
|
|||||||
//
|
//
|
||||||
|
|
||||||
struct ggml_object {
|
struct ggml_object {
|
||||||
size_t offs;
|
size_t offset;
|
||||||
size_t size;
|
size_t size;
|
||||||
|
|
||||||
struct ggml_object * next;
|
struct ggml_object * next;
|
||||||
@ -1284,9 +1277,6 @@ struct ggml_context {
|
|||||||
|
|
||||||
struct ggml_object * objects_begin;
|
struct ggml_object * objects_begin;
|
||||||
struct ggml_object * objects_end;
|
struct ggml_object * objects_end;
|
||||||
|
|
||||||
struct ggml_scratch scratch;
|
|
||||||
struct ggml_scratch scratch_save;
|
|
||||||
};
|
};
|
||||||
|
|
||||||
struct ggml_context_container {
|
struct ggml_context_container {
|
||||||
@ -1349,7 +1339,7 @@ inline static void ggml_critical_section_end(void) {
|
|||||||
|
|
||||||
void ggml_print_object(const struct ggml_object * obj) {
|
void ggml_print_object(const struct ggml_object * obj) {
|
||||||
GGML_PRINT(" - ggml_object: offset = %zu, size = %zu, next = %p\n",
|
GGML_PRINT(" - ggml_object: offset = %zu, size = %zu, next = %p\n",
|
||||||
obj->offs, obj->size, (const void *) obj->next);
|
obj->offset, obj->size, (const void *) obj->next);
|
||||||
}
|
}
|
||||||
|
|
||||||
void ggml_print_objects(const struct ggml_context * ctx) {
|
void ggml_print_objects(const struct ggml_context * ctx) {
|
||||||
@ -1545,14 +1535,12 @@ struct ggml_context * ggml_init(struct ggml_init_params params) {
|
|||||||
}
|
}
|
||||||
|
|
||||||
*ctx = (struct ggml_context) {
|
*ctx = (struct ggml_context) {
|
||||||
/*.mem_size =*/ params.mem_size,
|
.mem_size = params.mem_size,
|
||||||
/*.mem_buffer =*/ params.mem_buffer ? params.mem_buffer : malloc(params.mem_size),
|
.mem_buffer = params.mem_buffer ? params.mem_buffer : malloc(params.mem_size),
|
||||||
/*.mem_buffer_owned =*/ params.mem_buffer ? false : true,
|
.mem_buffer_owned = params.mem_buffer ? false : true,
|
||||||
/*.n_objects =*/ 0,
|
.n_objects = 0,
|
||||||
/*.objects_begin =*/ NULL,
|
.objects_begin = NULL,
|
||||||
/*.objects_end =*/ NULL,
|
.objects_end = NULL,
|
||||||
/*.scratch =*/ { 0, 0, NULL, },
|
|
||||||
/*.scratch_save =*/ { 0, 0, NULL, },
|
|
||||||
};
|
};
|
||||||
|
|
||||||
ggml_assert_aligned(ctx->mem_buffer);
|
ggml_assert_aligned(ctx->mem_buffer);
|
||||||
@ -1575,7 +1563,7 @@ void ggml_free(struct ggml_context * ctx) {
|
|||||||
g_state.contexts[i].used = false;
|
g_state.contexts[i].used = false;
|
||||||
|
|
||||||
GGML_PRINT_DEBUG("%s: context %d with %d objects has been freed. memory used = %zu\n",
|
GGML_PRINT_DEBUG("%s: context %d with %d objects has been freed. memory used = %zu\n",
|
||||||
__func__, i, ctx->n_objects, ctx->objects_end->offs + ctx->objects_end->size);
|
__func__, i, ctx->n_objects, ctx->objects_end->offset + ctx->objects_end->size);
|
||||||
|
|
||||||
if (ctx->mem_buffer_owned) {
|
if (ctx->mem_buffer_owned) {
|
||||||
free(ctx->mem_buffer);
|
free(ctx->mem_buffer);
|
||||||
@ -1594,15 +1582,7 @@ void ggml_free(struct ggml_context * ctx) {
|
|||||||
}
|
}
|
||||||
|
|
||||||
size_t ggml_used_mem(const struct ggml_context * ctx) {
|
size_t ggml_used_mem(const struct ggml_context * ctx) {
|
||||||
return ctx->objects_end->offs + ctx->objects_end->size;
|
return ctx->objects_end->offset + ctx->objects_end->size;
|
||||||
}
|
|
||||||
|
|
||||||
size_t ggml_set_scratch(struct ggml_context * ctx, struct ggml_scratch scratch) {
|
|
||||||
const size_t result = ctx->scratch.data ? ctx->scratch.offs : 0;
|
|
||||||
|
|
||||||
ctx->scratch = scratch;
|
|
||||||
|
|
||||||
return result;
|
|
||||||
}
|
}
|
||||||
|
|
||||||
////////////////////////////////////////////////////////////////////////////////
|
////////////////////////////////////////////////////////////////////////////////
|
||||||
@ -1616,9 +1596,9 @@ struct ggml_tensor * ggml_new_tensor_impl(
|
|||||||
// always insert objects at the end of the context's memory pool
|
// always insert objects at the end of the context's memory pool
|
||||||
struct ggml_object * obj_cur = ctx->objects_end;
|
struct ggml_object * obj_cur = ctx->objects_end;
|
||||||
|
|
||||||
const size_t cur_offs = obj_cur == NULL ? 0 : obj_cur->offs;
|
const size_t cur_offset = obj_cur == NULL ? 0 : obj_cur->offset;
|
||||||
const size_t cur_size = obj_cur == NULL ? 0 : obj_cur->size;
|
const size_t cur_size = obj_cur == NULL ? 0 : obj_cur->size;
|
||||||
const size_t cur_end = cur_offs + cur_size;
|
const size_t cur_end = cur_offset + cur_size;
|
||||||
|
|
||||||
size_t size_needed = 0;
|
size_t size_needed = 0;
|
||||||
|
|
||||||
@ -1629,52 +1609,25 @@ struct ggml_tensor * ggml_new_tensor_impl(
|
|||||||
}
|
}
|
||||||
// align to GGML_MEM_ALIGN
|
// align to GGML_MEM_ALIGN
|
||||||
size_needed = ((size_needed + GGML_MEM_ALIGN - 1)/GGML_MEM_ALIGN)*GGML_MEM_ALIGN;
|
size_needed = ((size_needed + GGML_MEM_ALIGN - 1)/GGML_MEM_ALIGN)*GGML_MEM_ALIGN;
|
||||||
|
|
||||||
}
|
}
|
||||||
|
|
||||||
char * const mem_buffer = ctx->mem_buffer;
|
|
||||||
struct ggml_object * const obj_new = (struct ggml_object *)(mem_buffer + cur_end);
|
|
||||||
|
|
||||||
if (ctx->scratch.data == NULL || data != NULL) {
|
|
||||||
size_needed += sizeof(struct ggml_tensor);
|
size_needed += sizeof(struct ggml_tensor);
|
||||||
|
|
||||||
if (cur_end + size_needed + GGML_OBJECT_SIZE > ctx->mem_size) {
|
if (cur_end + size_needed + GGML_OBJECT_SIZE > ctx->mem_size) {
|
||||||
GGML_PRINT("%s: not enough space in the context's memory pool (needed %zu, available %zu)\n",
|
GGML_PRINT("%s: not enough space in the context's memory pool\n", __func__);
|
||||||
__func__, cur_end + size_needed + GGML_OBJECT_SIZE, ctx->mem_size);
|
|
||||||
assert(false);
|
assert(false);
|
||||||
return NULL;
|
return NULL;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
char * const mem_buffer = ctx->mem_buffer;
|
||||||
|
|
||||||
|
struct ggml_object * const obj_new = (struct ggml_object *)(mem_buffer + cur_end);
|
||||||
|
|
||||||
*obj_new = (struct ggml_object) {
|
*obj_new = (struct ggml_object) {
|
||||||
.offs = cur_end + GGML_OBJECT_SIZE,
|
.offset = cur_end + GGML_OBJECT_SIZE,
|
||||||
.size = size_needed,
|
.size = size_needed,
|
||||||
.next = NULL,
|
.next = NULL,
|
||||||
};
|
};
|
||||||
} else {
|
|
||||||
if (ctx->scratch.offs + size_needed > ctx->scratch.size) {
|
|
||||||
GGML_PRINT("%s: not enough space in the scratch memory\n", __func__);
|
|
||||||
assert(false);
|
|
||||||
return NULL;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (cur_end + sizeof(struct ggml_tensor) + GGML_OBJECT_SIZE > ctx->mem_size) {
|
|
||||||
GGML_PRINT("%s: not enough space in the context's memory pool (needed %zu, available %zu)\n",
|
|
||||||
__func__, cur_end + sizeof(struct ggml_tensor) + GGML_OBJECT_SIZE, ctx->mem_size);
|
|
||||||
assert(false);
|
|
||||||
return NULL;
|
|
||||||
}
|
|
||||||
|
|
||||||
data = (char * const) ctx->scratch.data + ctx->scratch.offs;
|
|
||||||
|
|
||||||
*obj_new = (struct ggml_object) {
|
|
||||||
.offs = cur_end + GGML_OBJECT_SIZE,
|
|
||||||
.size = sizeof(struct ggml_tensor),
|
|
||||||
.next = NULL,
|
|
||||||
};
|
|
||||||
|
|
||||||
//printf("scratch offs = %zu, size_needed = %zu\n", ctx->scratch.offs, size_needed);
|
|
||||||
|
|
||||||
ctx->scratch.offs += size_needed;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (obj_cur != NULL) {
|
if (obj_cur != NULL) {
|
||||||
obj_cur->next = obj_new;
|
obj_cur->next = obj_new;
|
||||||
@ -1685,9 +1638,9 @@ struct ggml_tensor * ggml_new_tensor_impl(
|
|||||||
|
|
||||||
ctx->objects_end = obj_new;
|
ctx->objects_end = obj_new;
|
||||||
|
|
||||||
//printf("%s: inserted new object at %zu, size = %zu\n", __func__, cur_end, obj_new->size);
|
//GGML_PRINT_DEBUG("%s: inserted new object at %zu\n", __func__, cur_end);
|
||||||
|
|
||||||
struct ggml_tensor * const result = (struct ggml_tensor *)(mem_buffer + obj_new->offs);
|
struct ggml_tensor * const result = (struct ggml_tensor *)(mem_buffer + obj_new->offset);
|
||||||
|
|
||||||
ggml_assert_aligned(result);
|
ggml_assert_aligned(result);
|
||||||
|
|
||||||
@ -1772,26 +1725,16 @@ struct ggml_tensor * ggml_new_tensor_4d(
|
|||||||
}
|
}
|
||||||
|
|
||||||
struct ggml_tensor * ggml_new_i32(struct ggml_context * ctx, int32_t value) {
|
struct ggml_tensor * ggml_new_i32(struct ggml_context * ctx, int32_t value) {
|
||||||
ctx->scratch_save = ctx->scratch;
|
|
||||||
ctx->scratch.data = NULL;
|
|
||||||
|
|
||||||
struct ggml_tensor * result = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, 1);
|
struct ggml_tensor * result = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, 1);
|
||||||
|
|
||||||
ctx->scratch = ctx->scratch_save;
|
|
||||||
|
|
||||||
ggml_set_i32(result, value);
|
ggml_set_i32(result, value);
|
||||||
|
|
||||||
return result;
|
return result;
|
||||||
}
|
}
|
||||||
|
|
||||||
struct ggml_tensor * ggml_new_f32(struct ggml_context * ctx, float value) {
|
struct ggml_tensor * ggml_new_f32(struct ggml_context * ctx, float value) {
|
||||||
ctx->scratch_save = ctx->scratch;
|
|
||||||
ctx->scratch.data = NULL;
|
|
||||||
|
|
||||||
struct ggml_tensor * result = ggml_new_tensor_1d(ctx, GGML_TYPE_F32, 1);
|
struct ggml_tensor * result = ggml_new_tensor_1d(ctx, GGML_TYPE_F32, 1);
|
||||||
|
|
||||||
ctx->scratch = ctx->scratch_save;
|
|
||||||
|
|
||||||
ggml_set_f32(result, value);
|
ggml_set_f32(result, value);
|
||||||
|
|
||||||
return result;
|
return result;
|
||||||
@ -2400,7 +2343,7 @@ struct ggml_tensor * ggml_repeat(
|
|||||||
result->op = GGML_OP_REPEAT;
|
result->op = GGML_OP_REPEAT;
|
||||||
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
|
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
|
||||||
result->src0 = a;
|
result->src0 = a;
|
||||||
result->src1 = b;
|
result->src1 = NULL;
|
||||||
|
|
||||||
return result;
|
return result;
|
||||||
}
|
}
|
||||||
@ -3016,7 +2959,9 @@ struct ggml_tensor * ggml_diag_mask_inf(
|
|||||||
// TODO: when implement backward, fix this:
|
// TODO: when implement backward, fix this:
|
||||||
//struct ggml_tensor * result = inplace ? ggml_view_tensor(ctx, a) : ggml_dup_tensor(ctx, a);
|
//struct ggml_tensor * result = inplace ? ggml_view_tensor(ctx, a) : ggml_dup_tensor(ctx, a);
|
||||||
struct ggml_tensor * result = ggml_view_tensor(ctx, a);
|
struct ggml_tensor * result = ggml_view_tensor(ctx, a);
|
||||||
struct ggml_tensor * b = ggml_new_i32(ctx, n_past);
|
|
||||||
|
struct ggml_tensor * b = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, 1);
|
||||||
|
((int32_t *) b->data)[0] = n_past;
|
||||||
|
|
||||||
result->op = GGML_OP_DIAG_MASK_INF;
|
result->op = GGML_OP_DIAG_MASK_INF;
|
||||||
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
|
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
|
||||||
@ -4348,9 +4293,7 @@ static bool ggml_compute_forward_mul_mat_use_blas(
|
|||||||
const int ne1 = dst->ne[1];
|
const int ne1 = dst->ne[1];
|
||||||
|
|
||||||
// TODO: find the optimal values for these
|
// TODO: find the optimal values for these
|
||||||
if (ggml_is_contiguous(src0) && ggml_is_contiguous(src1) && (
|
if (ggml_is_contiguous(src0) && ggml_is_contiguous(src1) && ne0 >= 32 && ne1 >= 32 && ne10 >= 32) {
|
||||||
(ne0 >= 32 && ne1 >= 32 && ne10 >= 32)
|
|
||||||
)) {
|
|
||||||
//printf("BLAS: %d %d %d\n", ne0, ne1, ne10);
|
//printf("BLAS: %d %d %d\n", ne0, ne1, ne10);
|
||||||
return true;
|
return true;
|
||||||
}
|
}
|
||||||
@ -7339,9 +7282,6 @@ void ggml_graph_compute(struct ggml_context * ctx, struct ggml_cgraph * cgraph)
|
|||||||
node->n_tasks = 1; // TODO: this actually is doing nothing
|
node->n_tasks = 1; // TODO: this actually is doing nothing
|
||||||
// the threads are still spinning
|
// the threads are still spinning
|
||||||
cur = sizeof(float)*(node->src0->ne[0]*node->src0->ne[1]);
|
cur = sizeof(float)*(node->src0->ne[0]*node->src0->ne[1]);
|
||||||
//printf("src0: ne0 = %d, ne1 = %d, ne = %d\n", node->src0->ne[0], node->src0->ne[1], node->src0->ne[0]*node->src0->ne[1]);
|
|
||||||
//printf("src1: ne0 = %d, ne1 = %d, ne = %d\n", node->src1->ne[0], node->src1->ne[1], node->src1->ne[0]*node->src1->ne[1]);
|
|
||||||
//printf("cur = %zu\n", cur);
|
|
||||||
} else {
|
} else {
|
||||||
cur = sizeof(ggml_fp16_t)*ggml_nelements(node->src1);
|
cur = sizeof(ggml_fp16_t)*ggml_nelements(node->src1);
|
||||||
}
|
}
|
||||||
|
|||||||
9
ggml.h
9
ggml.h
@ -301,13 +301,6 @@ struct ggml_cgraph {
|
|||||||
int64_t perf_time_us;
|
int64_t perf_time_us;
|
||||||
};
|
};
|
||||||
|
|
||||||
// scratch buffer
|
|
||||||
struct ggml_scratch {
|
|
||||||
size_t offs;
|
|
||||||
size_t size;
|
|
||||||
void * data;
|
|
||||||
};
|
|
||||||
|
|
||||||
struct ggml_init_params {
|
struct ggml_init_params {
|
||||||
// memory pool
|
// memory pool
|
||||||
size_t mem_size; // bytes
|
size_t mem_size; // bytes
|
||||||
@ -334,8 +327,6 @@ void ggml_free(struct ggml_context * ctx);
|
|||||||
|
|
||||||
size_t ggml_used_mem(const struct ggml_context * ctx);
|
size_t ggml_used_mem(const struct ggml_context * ctx);
|
||||||
|
|
||||||
size_t ggml_set_scratch(struct ggml_context * ctx, struct ggml_scratch scratch);
|
|
||||||
|
|
||||||
struct ggml_tensor * ggml_new_tensor(
|
struct ggml_tensor * ggml_new_tensor(
|
||||||
struct ggml_context * ctx,
|
struct ggml_context * ctx,
|
||||||
enum ggml_type type,
|
enum ggml_type type,
|
||||||
|
|||||||
850
whisper.cpp
850
whisper.cpp
File diff suppressed because it is too large
Load Diff
15
whisper.h
15
whisper.h
@ -113,16 +113,6 @@ extern "C" {
|
|||||||
int n_samples,
|
int n_samples,
|
||||||
int n_threads);
|
int n_threads);
|
||||||
|
|
||||||
// Convert RAW PCM audio to log mel spectrogram but applies a Phase Vocoder to speed up the audio x2.
|
|
||||||
// The resulting spectrogram is stored inside the provided whisper context.
|
|
||||||
// Returns 0 on success
|
|
||||||
WHISPER_API int whisper_pcm_to_mel_phase_vocoder(
|
|
||||||
struct whisper_context* ctx,
|
|
||||||
const float* samples,
|
|
||||||
int n_samples,
|
|
||||||
int n_threads);
|
|
||||||
|
|
||||||
|
|
||||||
// This can be used to set a custom log mel spectrogram inside the provided whisper context.
|
// This can be used to set a custom log mel spectrogram inside the provided whisper context.
|
||||||
// Use this instead of whisper_pcm_to_mel() if you want to provide your own log mel spectrogram.
|
// Use this instead of whisper_pcm_to_mel() if you want to provide your own log mel spectrogram.
|
||||||
// n_mel must be 80
|
// n_mel must be 80
|
||||||
@ -267,7 +257,6 @@ extern "C" {
|
|||||||
float thold_pt; // timestamp token probability threshold (~0.01)
|
float thold_pt; // timestamp token probability threshold (~0.01)
|
||||||
float thold_ptsum; // timestamp token sum probability threshold (~0.01)
|
float thold_ptsum; // timestamp token sum probability threshold (~0.01)
|
||||||
int max_len; // max segment length in characters
|
int max_len; // max segment length in characters
|
||||||
bool split_on_word; // split on word rather than on token (when used with max_len)
|
|
||||||
int max_tokens; // max tokens per segment (0 = no limit)
|
int max_tokens; // max tokens per segment (0 = no limit)
|
||||||
|
|
||||||
// [EXPERIMENTAL] speed-up techniques
|
// [EXPERIMENTAL] speed-up techniques
|
||||||
@ -285,7 +274,6 @@ extern "C" {
|
|||||||
|
|
||||||
// common decoding parameters:
|
// common decoding parameters:
|
||||||
bool suppress_blank; // ref: https://github.com/openai/whisper/blob/f82bc59f5ea234d4b97fb2860842ed38519f7e65/whisper/decoding.py#L89
|
bool suppress_blank; // ref: https://github.com/openai/whisper/blob/f82bc59f5ea234d4b97fb2860842ed38519f7e65/whisper/decoding.py#L89
|
||||||
bool suppress_non_speech_tokens; // ref: https://github.com/openai/whisper/blob/7858aa9c08d98f75575035ecd6481f462d66ca27/whisper/tokenizer.py#L224-L253
|
|
||||||
|
|
||||||
float temperature; // initial decoding temperature, ref: https://ai.stackexchange.com/a/32478
|
float temperature; // initial decoding temperature, ref: https://ai.stackexchange.com/a/32478
|
||||||
float max_initial_ts; // ref: https://github.com/openai/whisper/blob/f82bc59f5ea234d4b97fb2860842ed38519f7e65/whisper/decoding.py#L97
|
float max_initial_ts; // ref: https://github.com/openai/whisper/blob/f82bc59f5ea234d4b97fb2860842ed38519f7e65/whisper/decoding.py#L97
|
||||||
@ -341,9 +329,6 @@ extern "C" {
|
|||||||
// A segment can be a few words, a sentence, or even a paragraph.
|
// A segment can be a few words, a sentence, or even a paragraph.
|
||||||
WHISPER_API int whisper_full_n_segments(struct whisper_context * ctx);
|
WHISPER_API int whisper_full_n_segments(struct whisper_context * ctx);
|
||||||
|
|
||||||
// Language id associated with the current context
|
|
||||||
WHISPER_API int whisper_full_lang_id(struct whisper_context * ctx);
|
|
||||||
|
|
||||||
// Get the start and end time of the specified segment.
|
// Get the start and end time of the specified segment.
|
||||||
WHISPER_API int64_t whisper_full_get_segment_t0(struct whisper_context * ctx, int i_segment);
|
WHISPER_API int64_t whisper_full_get_segment_t0(struct whisper_context * ctx, int i_segment);
|
||||||
WHISPER_API int64_t whisper_full_get_segment_t1(struct whisper_context * ctx, int i_segment);
|
WHISPER_API int64_t whisper_full_get_segment_t1(struct whisper_context * ctx, int i_segment);
|
||||||
|
|||||||
Reference in New Issue
Block a user