Commit Graph

37 Commits

Author SHA1 Message Date
Georgi Gerganov
e30c679928
whisper : reorganize source code + improve CMake (#2256)
* scripts : update sync [no ci]

* files : reorganize [no ci]

* sync : llama.cpp

* cmake : link math library

* cmake : build normal ggml library

* files : move headers to include

* objc : fix path to ggml-metal.h

* ci : fix WHISPER_CUDA -> GGML_CUDA

* scripts : sync LICENSE [no ci]
2024-06-26 19:34:09 +03:00
Georgi Gerganov
5181494e9f build : update make / cmake 2024-06-18 09:39:40 +03:00
Borislav Stanimirov
f760756078
minor: add CMakeSettings.json to gitignore (#2094) 2024-05-08 11:03:21 +03:00
Georgi Gerganov
1b553b9817
gitignore : .vimspector.json 2024-03-16 16:26:35 +02:00
Felix
eff3570f78
server : add a REST Whisper server example with OAI-like API (#1380)
* Add first draft of server

* Added json support and base funcs for server.cpp

* Add more user input via api-request

also some clean up

* Add reqest params and load post function

Also some general clean up

* Remove unused function

* Add readme

* Add exception handlers

* Update examples/server/server.cpp

* make : add server target

* Add magic curl syntax

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2023-11-20 21:40:24 +02:00
Tong Li
54a08bde29
examples : add whisper.android.java for compatibility with older Android versions using Java (#1382)
* save the recorded audio to a file

* Alignment -help

* Save the correct audio

* chage to a consistent coding style

* Correct typo

* Update examples/stream/stream.cpp

* Update examples/stream/stream.cpp

* Correct variable misuse

* Update examples/stream/stream.cpp

* Update examples/stream/stream.cpp

* Update examples/stream/stream.cpp

* Update examples/stream/stream.cpp

* add *.bin .cxx/ .gradle/ cmake-build-debug/ to gitignore

* add whisper.android.java

* Added support for older versions of Android of Java

* add examples for android java

* add README.md for android java

* add fullTranscribeWithTime

* 增加 toString()方法和测试

* change return type to void

* update to v1.4.1

* add WhisperService

* chage to whisper_full_get_segment_t1

* add method transcribeDataWithTime

* modified toString
```
return "[" + start + " --> " + end + "]:" + sentence;
```

* Optimize code logic

* update text view on handle

* set max lines

* change Chinese to English

* Update bindings/java/build.gradle

* Update .gitignore

* add android.java to github action

* chage android.java to   android_java in build.yml

* remove gradle

* chage jdk to temurin in android_java of CI

* chage jdk to temurin 11 in android_java of CI

* add x to gradlew

* set api-level for android_java of CI

* Update examples/whisper.android.java/app/src/main/jni/whisper/CMakeLists.txt

* add ndk version in build.gradle

* remove local.properties

* add testFullTranscribeWithTime

---------

Co-authored-by: litongmacos <litongjava@qq.com>
Co-authored-by: bobqianic <129547291+bobqianic@users.noreply.github.com>
2023-11-12 18:31:58 +02:00
Georgi Gerganov
b0502836b8
whisper : add full CUDA and Metal offloading (#1472)
* whisper : migrate to ggml-backend

* whisper : fix logit reading

* whisper : fix tensor allocation during load

* whisper : fix beam-search with CUDA

* whisper : free backends + fix compile warning

* whisper : print when CUDA is enabled

* whisper : fix CoreML

* make : clean-up

* talk : fix compile warning

* whisper : support ggml_conv with CUDA and Metal (#1473)

* ggml : add CUDA support for ggml_conv

* whisper : remove ggml_repeat for conv bias + single backend

* cuda : fix im2col kernel

* metal : add im2col support + mul mat-vec f16 x f16

* bench-all : add q4 models

* whisper : clean-up

* quantize-all : fix

* ggml : im2col opts

* whisper : avoid whisper_model_data wrapper

* whisper : add note that ggml_mul_mat_pad does not work with CUDA

* whisper : factor out graph compute in common function

* whisper : fixes

* whisper : fix UB with measure buffers

* whisper : try to fix the parallel whisper_state functionality (#1479)

* whisper : try to fix the parallel whisper_state functionality

* whisper : fix multi-state Metal

* whisper : free backend instances in whisper_state
2023-11-12 15:31:08 +02:00
Sindre Sorhus
d03c60dd7f
ios : add support for Swift Package Manager (#1370)
* Add support for Swift

* Make it build in Xcode

* Use the SPM package in the SwiftUI example app
2023-11-07 23:53:31 +02:00
Neil Chudleigh
9edbd0a204
extra: Add benchmark script implemented in Python (#1298)
* Create bench.py

* Various benchmark results

* Update benchmark script with hardware name, and file checks

* Remove old benchmark results

* Add git shorthash

* Round to 2 digits on calculated floats

* Fix the header reference when sorting results

* FIx order of models

* Parse file name

* Simplify filecheck

* Improve print run print statement

* Use simplified model name

* Update benchmark_results.csv

* Process single or lists of processors and threads

* Ignore benchmark results, dont check in

* Move bench.py to extra folder

* Readme section on how to use

* Move command to correct location

* Use separate list for models that exist

* Handle subprocess error in git short hash check

* Fix filtered models list initialization
2023-09-25 23:45:15 +08:00
Georgi Gerganov
0839209cab
gitignore : update 2023-09-08 19:45:28 +03:00
AustinMroz
175ffa64ee
examples : vim plugin and LSP server (#1144)
* Initial proof of concept Vim plugin

At present, this is likely only slightly better than feature parity with
the existing whisper.nvim

Known issues:
 Trailing whitespace
 Up to an existing length(5 seconds) of speech may be processed when
  listening is enabled
 CPU cycles are spent processing speech even when not listening.

Fixing these issues is likely dependent upon future efforts to create a
dedicated library instead of wrapping examples/stream

* Support $WHISPER_CPP_HOME environment variable

A minor misunderstanding of the whisper.nvim implementation resulted in
a plugin that was functional, but not a drop in replacement as it should
be now.

* Initial progress on LSP implementation

Libcall is nonviable because the library is immediately freed after a
call is made. Further investigation has shown Language Server Protocol as
a promising alternative that both simplifies the required logic on the
vimscript side and increases the ease with which plugins for other
editors could be made in the future. This is a very large undertaking
and my progress has slowed substantially.

Work is far from being in a usable state, but I wish to keep track of
major refactors for organizational purposes.

* Rewrite audio windowing of guided transcription

One of the defining goals of this venture is allowing consecutive
commands to be rattled off without the existing deadzones of the current
implementation.

* Add unguided_transcription. Cleanup.

The unguided transcription implantation heavily borrows from existing
example implementations and the guided_transcription logic.

A high level pass was done to check that method arguments are accurate
to what inputs are actually required.

A first attempt at cancellation support was added for record keeping,
but will be deleted in a future commit.

* Fix compilation.

Resolves a large number of compilation errors.
No testing has been done yet for execution errors.

Update Makefile and .gitignore

* Functional unguided_transcription

* Functional guided_transcription

Fix commandset_list being passed by value
Properly register the first token of a multitoken command

* Minor changes before time fix

I've apparently made an awfully major mistake in thinking that unix time
was in milliseconds and will be changing all timekeeping code to use
standardized methods.

In preparation for this is a number of minor bugfixes.
Output is manually flushed.
An echo method has been added.
registerCommandset now wraps the returned index

* Swap timekeeping to use std::chrono

* Add work in progress lsp backed whisper.vim plugin

Current progress blockers are
 Adding modality awareness to the command processing
  (specifically, motion prompting)
 Improving the VAD to be a little more responsive
  (testing start of activity)

* Reworked vim plugin command loop

* Fix change inside

Multiple bug fixes that, crucially, bring the plugin to the point where a
demonstration video is possible

Add better echo messaging so whisper_log isn't required
 Add loading complete message as indicator when listening has started
Insert/append are actually included in command sets
Some more heavy handed corrections to prevent a double exit when leaving
insert mode
As a somewhat hacky fix, the very first space is removed when inserting.
 This cleans up most use cases, but leaves me unsatisfied with the few
 cases it would be desired.

* Forcibly set commandset_index to 0 after subinsert

Also remove unnecessary ! to use builtin vim command

* Fix upper

A minor scope mistake was causing upper'd inputs to be eaten.
This was fixed and echoing was slightly improved for clarity.

* Fix formatting

Corrects indentation to 4 spaces as project standard
Slightly better error support for malformed json input

* Remove obsolete vim plugin

* Add json.hpp library

The same library that is used for the llama.cpp server

* Minor cleanups

add lsp to the make clean directive.
remove a redundant params definition.
reorder whisper.vim logging for subtranscriptions
Corrections to unlets (variables of argument scope appear immutable)

* Fix indentation. Fallback for subTranscription

Indentation has been changed to 4 spaces.

Unit testing has been set up, I'm opting not to include it in the
repository for now.
It however, has revealed a bug in the state logic where a
subtranscription can be initiated without having a saved command
When this occurs, append is added as a fallback

* Move audio polling logic to a subfunction

While work on the improved vad will continue, It's grown to be a little
out of scope. Instead, a future commit will perform multiple detection
passes at substretches of audio when a backlog of audio exists.

To facilitate this, and prevent code duplication, the vad code has been
moved into a subfunction shared by both the unguided and guided
transcription functions.

* Test for voice over subchunks if backlog > 1s

As the existing VAD implementation only checks for a falling edge at the
end of an audio chunk. It fails to detect voice in cases where the
recorded voice is only at the beginning of the audio.

To ameliorate this, when the timestamp would cause analysis of audio
over a second in length, it is split into 1 second length subchunks
which are individually tested.

Results are promising, but there seems to be a remaining bug with
unguided transcription likely related to saving context

* Limit the maximum length of audio input.

This existing VAD implementation only detects falling edges, which
means any gap in the users speaking is processed for transcription.
This simply establishes a constant maximum length depending on the type
of transcription. Uguided gets a generous 10 seconds and guided, 2.

While quick testing showed that commands are generally around a half a
second to a second, limiting commands to an even second resulted in
extreme degradation of quality. (Seemingly always the same output for a
given commandset)

* Unguided timestamp tracking, cleanup

Unguided transcriptions where not setup to allow for passing of
timestamp data forward, but have been corrected.

No_context is now always set to false. While conceptually desirable for
the quality of guided transcription, It was seemingly responsible for
prior command inputs ghosting in unguided transcription.

Save and Run are now tracked by command number instead of command text.
While command_text was provided for convenience, I wish to keep command
index authoritative. This gives greater consistency and potentially
allows for end users to rename or even translate the spoken versions of
these commands

* By default, maintain mode.

Previously, mode was reset to 0 unless otherwise set.
In addition to causing some edge cases, this was didn't mesh well with
the existing approach to visual mode.

With this change, initial tests indicate visual mode is functional.

* Add undo breaks before subtranscriptions

Subtranscriptions use undo as a hack to allow for partial responses to
be displayed. However, scripts don't cause an undo break mid execution
unless specifically instructed to. This meant that multiple
unguided transcriptions from a single session would cause a latter to
undo a former.

This is now fixed and undo should be reasonably usable as a command.

* Append instead of insert for new undo sequence

When entering and leavening insert mode with `i`, the cursor shifts one
column to the left. This is remedied by using append instead of insert
for setting these breaks in the undo sequence

`-` was also added to the pronunciation dictionary to be pronounced as
minus as it was causing a particularly high failure rate.

* Move undo sequence breaks to command execution

Previously, undo sequence breaks were triggered when there was a command
that caused a move to insert mode. This caused commands that changed
state (like delete or paste) to be bundled together with into the last
command that caused text to be entered.

* Fix repeat. Add space, carrot, dollar commands

 Repeat (.) wasn't being tracked properly just like undo and is being
 manually tracked now.

 While efforts have been made to properly handle spaces, it was
 particularly finicky to add a single space when one is needed. A
 special 'space' command has been added to insert a single space and move
 the cursor after it.

 Carrot and Dollar commands have been added for start of line and end of
 line respectively. These are both simple to implement, and just a
 matter of defining a pronunciation.

* Return error on duplicate in commandset

Not every command in the commandset tokenizes to a single token.
Because of this, it's possible for that two commands could resolve to
the same single token after subsequent tokens are discarded.

This commit adds a simple check for duplicates when a commandset is
registered and returns an error if so.

Additional code will be required later on the vim side to actually
process this error.

* Add support for user-defined commands

This adds a user definable dictionary from spoken keys to strings or
funcrefs. All keys are added to the commandlist and when spoken, trigger
the corresponding function.

Like "save" and "run", these user commands are only available when the
command buffer is empty.

* Add readme, update cmake

* Add area commandset. Refactor spoken_dict

Area commands (inside word, around sentence...) have been given a
commandset as considered earlier.

Verbose definitions for spoken_dict entries now use dicts instead of
lists. This shortens the definition for most keys that require it and
scales better with the addition of further commandsets

* Add mark, jump. Fix change under visual.

Mark (m) and jump (') have been added.

When a visual selection was executed upon a command that initiated a
subtranscription (change) the area of the visual selection is not
properly tracked which causes the attempt to stream in partial response
to fail. This is solved by disabling partial transcriptions from being
streamed when a subtranscription is started while in visual mode.

* Accommodate ignorecase. Fix change.

From testing on older different versions of vim, the test for
distinguishing an 'R' replace all from an 'r' replace could fail if
ignorecase was set. The comparison has been changed to explicitly
require case matching

Change detection has been moved to the execution section as it was missing the
change+motion case.

* Support registers. Fix README typo

There's no logic to prevent doubled register entry, but the functional
result is equivalent to if the same key order was typed into vim.

A minor typo in the readme. I've mismemorized the mnemonic for 't' as 'to'
instead of till., but 'to' can't be used as it's a homophone with '2'.
While there was no mistake in the actual logic, it was misleading to use
'to' in the readme.
2023-08-27 21:35:06 +03:00
Alexey Kharlamov
041be06d58
cmake : build with any BLAS compatible library (#927)
* Build with any BLAS library

* ci: Removed explicit CUDA nvcc path
2023-05-20 21:23:45 +03:00
Nicholas Albion
bc89f285d8
bindings : add java bindings (#931)
* WIP - java bindings

* updated README

* failed attempt at JNI

* fullTranscribe() test passes

* tested on Ubuntu 20

* link to Java bindings
2023-05-20 18:25:02 +03:00
Georgi Gerganov
794b162a46
whisper : add integer quantization support (#540)
* whisper : add integer quantization support

* examples : add common-ggml + prepare to add "quantize" tool

* whisper : quantization tool ready

* whisper : fix F32 support

* whisper : try to fix shared lib linkage

* wasm : update quantized models to Q5

* bench.wasm : remove "medium" button

* bench.wasm : fix custom model button

* ggml : add Q5_0 and Q5_1 WASM SIMD

* wasm : add quantized models to all WASM examples

* wasm : bump DB version number to 2

* talk-llama : update example to latest llama.cpp

* node : increase test timeout to 10s

* readme : add information for model quantization

* wasm : add links to other examples
2023-04-30 18:51:57 +03:00
Georgi Gerganov
5fd1bdd7fc
whisper : add GPU support via cuBLAS (#834)
* make : add WHISPER_CUBLAS

* make : fix CUBLAS build

* whisper : disable Flash Attention + adjust memory buffers

* whisper : remove old commented code

* readme : add cuBLAS instructions

* cmake : add WHISPER_CUBLAS option

* gitignore : ignore build-cublas
2023-04-30 12:14:33 +03:00
Georgi Gerganov
5e47e223bd
whisper : add Core ML support (#566)
* coreml : use Core ML encoder inference

* coreml : simlpify whisper_encode + log messages

* whisper : resolve rebase conflicts

* coreml : add scripts for CoreML model generation

* bench-all : recognize COREML flag
2023-04-15 13:21:27 +03:00
Georgi Gerganov
34b772727d
gitignore : add .test 2023-04-14 20:13:47 +03:00
Georgi Gerganov
4a0deb8b1e
talk-llama : add new example + sync ggml from llama.cpp (#664)
* talk-llama : talk with LLaMA AI

* talk.llama : disable EOS token

* talk-llama : add README instructions

* ggml : fix build in debug
2023-03-27 21:00:32 +03:00
Georgi Gerganov
ad1389003d
release : v1.2.1 2023-02-28 22:29:12 +02:00
Lukas Rist
02c7516c57
go : added wrappers to reset and print timings (#436) 2023-01-25 18:57:30 +02:00
Georgi Gerganov
9a65269a20
.gitignore : add arm_neon.h 2023-01-23 20:19:04 +02:00
Georgi Gerganov
8de452c18b
Improve decoding (#291)
* whisper : prepare infra for new decoding strategies

* whisper : apply logit filters and compute logprobs

* whisper : add whisper_get_logits()

* whisper : separate self and cross attention memory

Initial step needed for supporting parallel decoders

* whisper : move probs_id buffer to whisper_context

* whisper : refactor kv cache into separate struct

* whisper : move self-attention kv cache to whisper_decoder

* whisper : wip decoding parameters + strategies

* whisper : wip decoding parameters + strategies (part 2)

* whisper : wip decoding parameters + strategies (part 3)

* whisper : wip decoding parameters + strategies (part 4)

* whisper : fix prompt_past update to not include prompt_init

* whisper : temperature + best_of support

* whisper : support for compression_ration_threshold

We actually use entropy, but it is similar

* command : fix example to use logits instead of obsolete probs

* whisper : handle empty sequence ranking

* whisper : add WHISPER_DEBUG + diagnostic prints + new main args

* whisper : minor fixes

* whisper : add beam-search support

* whisper : bug fix when there no previous context

* whisper : add comments

* stream : disable temperature fallback

For real-time processing, we always want a single decoder running at T=0

* whisper.swiftui : update example - fix paths + add empty folders
2023-01-15 11:29:57 +02:00
Georgi Gerganov
054940e1f6
minor : fix .gitignore to not ignore examples 2022-12-11 11:39:46 +02:00
Georgi Gerganov
3b1aacbe6d talk : talk with AI in the terminal 2022-12-10 16:51:58 +02:00
Georgi Gerganov
832b4f34c9
make : indentation + .gitignore 2022-12-08 19:42:06 +02:00
Georgi Gerganov
bc88eb13c6
examples : add "command" tool (#171) 2022-11-25 19:36:57 +02:00
Georgi Gerganov
b8ce25dec1
refactoring : more readable code 2022-11-25 19:28:04 +02:00
Georgi Gerganov
c6710efde2 refactoring : move main + stream in examples + other stuff 2022-10-25 20:53:48 +03:00
Georgi Gerganov
bb1ee266d2 ios : whisper.objc example 2022-10-24 18:23:07 +03:00
Georgi Gerganov
d6b84b2a23
ref #62 : fix build for some compilers
For some reason, new version of GCC panic when the struct type is not
specified explicitly
2022-10-18 10:57:03 +03:00
Georgi Gerganov
b4a3875b2c
Revert recent sampling change
It does not actually help and seems to produce worse results on some of
the samples
2022-10-18 08:26:16 +03:00
Georgi Gerganov
0e858f080d
close #56 : build on FreeBSD
Thanks to @abelbabel for the contribution
2022-10-17 18:10:16 +03:00
Borislav Stanimirov
28252352d7 Visual Studio ignored dirs 2022-10-11 20:57:33 +03:00
Georgi Gerganov
2f069335ab Adding sanitizer tests 2022-10-08 11:43:42 +03:00
Georgi Gerganov
877c058179 Add CMake support 2022-10-08 09:02:41 +03:00
Georgi Gerganov
b6bf906730
ref #10 : quick-and-dirty attempt for real-time audio transciption
- Processes input in chunks of 3 seconds.
- Padding audio with silence
- Uses 1 second audio from previous pass
- No text context
2022-10-02 17:55:45 +03:00
Georgi Gerganov
b0a11594ae
Initial release 2022-09-25 22:13:49 +03:00