guillaumekln/faster-whisper issues and pull requests

#451 - help Unknown cover type: 0x1.

Issue - State: open - Opened by lilinrestart about 1 year ago - 5 comments

#449 - Expose generation parameter `no_repeat_ngram_size`

Pull Request - State: closed - Opened by guillaumekln about 1 year ago

#448 - Force the garbage collector to run after decoding the audio with PyAV

Pull Request - State: closed - Opened by guillaumekln about 1 year ago

#447 - App crash on low VRAM instead of "out of memory error"

Issue - State: open - Opened by Purfview about 1 year ago - 3 comments

#446 - AttributeError: module 'collections' has no attribute 'MutableMapping'

Issue - State: closed - Opened by pontifex03 about 1 year ago

#445 - Added audio duration after VAD to TranscriptionInfo object

Pull Request - State: closed - Opened by MinorJinx about 1 year ago - 3 comments

#444 - Get alternatives for each segment

Issue - State: open - Opened by Ar770 about 1 year ago - 2 comments

#443 - Can I get phoneme-level and its confidence score?

Issue - State: closed - Opened by tiennguyen12g about 1 year ago - 4 comments

#442 - (WSL2) RuntimeError: CUDA failed with error out of memory

Issue - State: open - Opened by coder543 about 1 year ago - 16 comments

#441 - Performance Degradation When Running Multiple Jobs in Parallel on a Single GPU

Issue - State: open - Opened by recog-arch about 1 year ago - 5 comments

#440 - int8 in CTranslate2 v3.18.0 [issues with int8_float16]

Issue - State: closed - Opened by Purfview about 1 year ago - 4 comments

#438 - Initial Prompt length + Execution time

Issue - State: closed - Opened by Arasun01 about 1 year ago - 1 comment

#437 - Install Faster Whisper off line

Issue - State: closed - Opened by Dvk2002 about 1 year ago - 5 comments

#436 - Repeated Word Issue In Silent Parts

Issue - State: open - Opened by makoto-toyouke about 1 year ago - 4 comments

#435 - Some fragments recognize no output

Issue - State: open - Opened by wdxiaobai about 1 year ago - 2 comments

#433 - question: related to faster-whisper installation.

Issue - State: closed - Opened by AbubakrChan about 1 year ago - 1 comment

#432 - When using faster-whisper, how to automatically split sentences

Issue - State: closed - Opened by OlalalalaO about 1 year ago - 2 comments

#430 - Can I run t as c++ application?

Issue - State: closed - Opened by Tonku about 1 year ago - 2 comments

#429 - ProxyError SSLError SSLEOFError: EOF occurred in violation of protocol

Issue - State: closed - Opened by Purfview over 1 year ago - 4 comments

#428 - Is it possibile to controll the number of seconds the input audio file is chuncked?

Issue - State: closed - Opened by Lorenzoncina over 1 year ago - 5 comments

#427 - docs: add wscribe to community integrations

Pull Request - State: closed - Opened by geekodour over 1 year ago

#426 - Faster-whisper model not working with torchserve inside docker container

Issue - State: closed - Opened by suryansh-noice over 1 year ago - 6 comments

#425 - Request to update the benchmark table

Issue - State: open - Opened by csukuangfj over 1 year ago - 4 comments

#424 - Faster Whisper large-v2 model is repeating the segments

Issue - State: open - Opened by Talhazeb over 1 year ago - 16 comments

#423 - Why doesn't faster-whisper require ffmpeg？

Issue - State: closed - Opened by weiyechaocat over 1 year ago - 1 comment

#422 - Why doesn't fast wheeler require ffmpeg

Issue - State: closed - Opened by weiyechaocat over 1 year ago

#420 - Process finished with exit code -1073740791 (0xC0000409)

Issue - State: closed - Opened by BazzaRB over 1 year ago - 1 comment

#419 - FeatureExtractor consumes a lot of cpu resources

Issue - State: open - Opened by Emma-Miao over 1 year ago - 5 comments

#418 - Unexpected behavior if a real language differs from the parameter.

Issue - State: closed - Opened by exactstat over 1 year ago - 1 comment

#417 - Repetition penalty adjustable

Pull Request - State: closed - Opened by chainyo over 1 year ago

#415 - Question about compute evaluation metric

Issue - State: open - Opened by DuyTa506 over 1 year ago - 3 comments

#413 - Compression ratio threshold is not met

Issue - State: closed - Opened by vadi2 over 1 year ago - 6 comments

#410 - Some of the improvements in Improve timestamp heuristics don't seem to be applying

Issue - State: closed - Opened by archive-r over 1 year ago - 3 comments

#409 - Sharing some code using fast whisper + spacy + PunctuationModel. need some refinement.

Issue - State: closed - Opened by search620 over 1 year ago

#408 - RuntimeError: cuDNN failed with status CUDNN_STATUS_NOT_INITIALIZED

Issue - State: closed - Opened by silvacarl2 over 1 year ago - 1 comment

#407 - If can't decode audio stream from video file , raise tuple index error

Issue - State: closed - Opened by iorilu over 1 year ago - 3 comments

#405 - Why is not possible to use this tool offline?

Issue - State: closed - Opened by fatualux over 1 year ago - 3 comments

#404 - Is Faster-whisper support --print-colors

Issue - State: closed - Opened by tiennguyen12g over 1 year ago - 2 comments

#403 - added repetition_penalty to TranscriptionOptions

Pull Request - State: closed - Opened by bfs15 over 1 year ago

#402 - Request for GSM Audio Support

Issue - State: closed - Opened by dorlanpabon over 1 year ago - 2 comments

#399 - Adds DEBUG log message for prompt_reset_on_temperature

Pull Request - State: closed - Opened by Purfview over 1 year ago

#398 - Rename clear_previous_text_on_temperature argument

Pull Request - State: closed - Opened by Purfview over 1 year ago - 2 comments

#397 - Add clear_previous_text_on_temperature parameter

Pull Request - State: closed - Opened by archive-r over 1 year ago

#396 - Reduce startup/warmup time

Issue - State: closed - Opened by emcodem over 1 year ago - 2 comments

#395 - Add a threshold_limit_ratio parameter

Pull Request - State: closed - Opened by archive-r over 1 year ago - 6 comments

#394 - Token repeat limit, hallucination helper option.

Pull Request - State: closed - Opened by bfs15 over 1 year ago - 9 comments

#392 - What is beam size?

Issue - State: closed - Opened by vadi2 over 1 year ago - 3 comments

#391 - AttributeError: module 'ctranslate2' has no attribute 'StorageView'

Issue - State: closed - Opened by Geremia over 1 year ago - 2 comments

#390 - Memory Leak investigation

Issue - State: closed - Opened by adambuar over 1 year ago - 4 comments

#387 - Is there a way to work with Whisper models finetuned with transformers?

Issue - State: closed - Opened by Jeronymous over 1 year ago - 2 comments

#386 - Audio stream using web-sockets not able to be transcribed by faster-whisper.

Issue - State: closed - Opened by JayAdityaNautiyal over 1 year ago - 1 comment

#385 - fix subtitle start time advance or end time delay, when set vad_filter to True

Pull Request - State: closed - Opened by zhoushuo19 over 1 year ago

#384 - Streaming support

Issue - State: open - Opened by AmgadHasan over 1 year ago - 5 comments

#383 - Patching word alignment to account for VAD timestamps, as well as punctuation "durations"

Pull Request - State: closed - Opened by aleksandr-smechov over 1 year ago

#382 - Strange performance behaviour

Issue - State: closed - Opened by x86Gr over 1 year ago - 9 comments

#381 - How to push to replicate.com?

Issue - State: closed - Opened by mukhtharcm over 1 year ago - 1 comment

#380 - bfloat16 support

Issue - State: closed - Opened by Sharrnah over 1 year ago - 2 comments

#379 - bfloat16 and int8_bfloat16

Issue - State: closed - Opened by Purfview over 1 year ago - 3 comments

#378 - Kernel crash with Tesla P40 GPU on CUDA 12.1, but works fine on Google Colab with CUDA 12.0

Issue - State: closed - Opened by mabergerx over 1 year ago - 7 comments

#377 - "Processing segment at..." logger is incorrect when VAD is enabled

Issue - State: closed - Opened by Purfview over 1 year ago - 1 comment

#376 - Invalidate the cached encoder output when no_speech threshold is met

Pull Request - State: closed - Opened by guillaumekln over 1 year ago

#375 - File t does not exist in model openai/whisper-large-v2

Issue - State: closed - Opened by Ibuprofenighty over 1 year ago - 5 comments

#374 - model.transcribe takes really long time on a short (5 sec) mp3 file

Issue - State: closed - Opened by harelfar2 over 1 year ago - 10 comments

#373 - Strange error: "killed?"

Issue - State: closed - Opened by ckcollab over 1 year ago - 7 comments

#372 - Not enough values to unpack (expected 2, got 1)

Issue - State: closed - Opened by justKiri over 1 year ago

#369 - no transcription results

Issue - State: open - Opened by lukeewin over 1 year ago - 1 comment

#364 - VAD is relatively slow

Issue - State: open - Opened by AlexandderGorodetski over 1 year ago - 24 comments

#361 - When I retrieve audio data from the WebSocket client and process it with io.BytesIO(audio_bytes), it still throws an error.

Issue - State: closed - Opened by lukeewin over 1 year ago - 11 comments

#360 - Normalizer issue

Issue - State: closed - Opened by dipeshhoncho07 over 1 year ago - 3 comments

#354 - Task: translate hallucinations

Issue - State: closed - Opened by joseph2mi over 1 year ago - 9 comments

#351 - Missing ending of transcription

Issue - State: open - Opened by hayrapetyan8768 over 1 year ago - 7 comments

#350 - questions about quantification

Issue - State: closed - Opened by breaddance over 1 year ago - 3 comments

#345 - semaphore error

Issue - State: closed - Opened by Maxamed over 1 year ago - 4 comments

#338 - word_level timestamps are 00:00

Issue - State: open - Opened by Ar770 over 1 year ago - 13 comments

#336 - Improve timestamp heuristics

Pull Request - State: closed - Opened by archive-r over 1 year ago

#333 - Question: word candidates

Issue - State: closed - Opened by exactstat over 1 year ago - 3 comments

#323 - Multi-gpu leads to `RuntimeError: CUDA failed with error invalid argument`

Issue - State: closed - Opened by chainyo over 1 year ago - 11 comments

#322 - Model breaks during presence of silence + speech

Issue - State: closed - Opened by sahamrit over 1 year ago - 6 comments

#321 - Problem with model conversion

Issue - State: closed - Opened by boykan313 over 1 year ago - 7 comments

#319 - Interrupted Word and punctuations

Issue - State: closed - Opened by SunHaoRanCN over 1 year ago - 1 comment

#315 - faster-whisper is the same speed as openai/whisper for cpu & `beam_size=1`

Issue - State: closed - Opened by stasbel over 1 year ago - 8 comments

#295 - Automatic Inference Batching Support

Issue - State: open - Opened by SinanAkkoyun over 1 year ago - 3 comments

#294 - Word-level timestamps are very inaccurate

Issue - State: closed - Opened by a-rogalska over 1 year ago - 2 comments

#279 - With CPU, it's quite slow. A 2-minute MP3 takes 6 minutes to finish.

Issue - State: closed - Opened by zxl777 over 1 year ago - 5 comments

#261 - 504 Server Error: Gateway Time-out for url

Issue - State: closed - Opened by billyg88 over 1 year ago - 3 comments

#260 - Enable onnxruntime dependency for Python 3.11

Pull Request - State: closed - Opened by guillaumekln over 1 year ago

#259 - Make get_speech_timestamps backward compatible with the previous usage

Pull Request - State: closed - Opened by guillaumekln over 1 year ago

#258 - Export version at the module level

Pull Request - State: closed - Opened by guillaumekln over 1 year ago

GitHub / guillaumekln/faster-whisper issues and pull requests