Commit Graph

  • 8c3680cb02 Add requires_hpu marker in tests PiotrBLL 2024-10-31 13:54:53 +01:00
  • b7069e579d Add initial imports and usage of wrap_in_hpu_graph PiotrBLL 2024-10-31 13:54:29 +01:00
  • 9b8c3cbf2d Update setup.py to accommodate Python3.13.0 Jason Shipp 2024-10-26 17:45:30 -04:00
  • 5979f03701
    Add option to carry initial_prompt with the sliding window (#2343) kittsil 2024-10-26 09:17:31 -05:00
  • 1778952a2f
    Merge branch 'main' into patch-1 Jong Wook Kim 2024-10-26 07:14:49 -07:00
  • 4fbca552cb
    Merge ad84a5f266e4e274a0076fa8fcf98aec3ab7580f into cdb81479623391f0651f4f9175ad986e85777f31 Kent Slaney 2024-10-26 14:41:15 +02:00
  • a0d19f1557
    Merge branch 'main' into fix/torch-load-weights-only-warning Jong Wook Kim 2024-10-25 17:30:10 -07:00
  • cdb8147962
    more pytorch versions in tests (#2408) Jong Wook Kim 2024-10-25 17:30:02 -07:00
  • dc963eba3b
    more pytorch versions in tests jongwook/more-pytorch-versions Jong Wook Kim 2024-10-25 17:23:06 -07:00
  • eafc2ebd1d
    Create CUALI braylexsanchez 2024-10-25 10:47:58 -05:00
  • bdbe6bfb47 add TranscribeProgressReceiver for update monitoring Austin Chang 2024-10-19 11:20:55 +08:00
  • 38f4e0919b Merge branch 'main' into jongwook-patch-1 suppletester 2024-10-17 03:40:04 +02:00
  • 81ef7f3bad
    Improvements implemented in the audio processing module mrfelpa 2024-10-15 18:02:43 +00:00
  • 2aafd81de3
    repeats audio_features tensor, just like tokens tensor, by group size for beam search or best-of-n sampling, to avoid size mismatch error in attention by batch inference Ruhollah Majdoddin 2024-10-14 19:35:29 +02:00
  • 0390c216d6 fix mercury233 2024-10-10 11:38:06 +08:00
  • daf8ba8fe1 typo mercury233 2023-03-07 11:24:38 +08:00
  • 01768370d8 add always_use_initial_prompt mercury233 2023-03-07 11:22:03 +08:00
  • a9c001394a update get_writer mercury233 2023-04-15 11:50:27 +08:00
  • dc6036eab8 get writer before loading model mercury233 2023-03-07 10:01:23 +08:00
  • de14ad8729 don't force ArgumentParser output_format choice mercury233 2023-03-07 09:59:56 +08:00
  • 7f0b03461a support comma separated output_format mercury233 2023-03-07 09:59:13 +08:00
  • 9b314370f3
    Update README.md BotMaster3000 2024-10-07 16:14:27 +02:00
  • dd002e3321
    Create devcontainer.json rushGitz 2024-10-05 13:52:31 -04:00
  • 76db75f5b5 Fix for ffmpeg return code false positive errors Raivis Dejus 2024-10-04 11:24:13 +03:00
  • 34ead17d74
    Create Entrevistas Pablo1061 2024-10-03 13:20:43 -03:00
  • 089f08945b
    Merge branch 'main' into main Yuekai Zhang 2024-10-03 12:25:14 +08:00
  • 022b7aae8f
    Merge branch 'main' into main Ashish Patel 2024-10-02 21:16:51 +05:30
  • ad84a5f266
    Merge branch 'main' into main Kent Slaney 2024-10-02 08:27:39 -07:00
  • feb45ab8d8
    Merge branch 'main' into fix/torch-load-weights-only-warning Ultr4_dev 2024-10-02 13:04:15 +02:00
  • cb76906e06
    Merge 5ed89d0ca24e99e93376da4626690354c1f1c6e7 into 25639fc17ddc013d56c594bfbf7644f2185fad84 jax 2024-10-01 14:18:29 +05:30
  • 3c32f478d3
    Merge 3537c29f93f24a790dd606c8a1f47d02225e37fe into 25639fc17ddc013d56c594bfbf7644f2185fad84 Adam Gardner 2024-10-01 09:56:10 +02:00
  • fcc8375995
    Merge 20e323895d942143f7a69e403bcbcbc33074e123 into 25639fc17ddc013d56c594bfbf7644f2185fad84 Gustavo Garcia 2024-09-30 22:02:16 +03:00
  • 25639fc17d Release 20240930 v20240930 Jong Wook Kim 2024-09-30 11:20:53 -07:00
  • 260bbcfcb3
    allowing numpy 2 in tests (#2362) Jong Wook Kim 2024-09-30 11:18:17 -07:00
  • 1e8530765a allowing numpy 2 in tests jongwook/numpy2-in-test Jong Wook Kim 2024-09-30 11:13:48 -07:00
  • f68af3098c allowing numpy 2 in tests Jong Wook Kim 2024-09-30 11:11:54 -07:00
  • 25e5c364e0
    large-v3-turbo model (#2361) Jong Wook Kim 2024-09-30 10:59:51 -07:00
  • f8b142fdaa large-v3-turbo model jongwook/large-v3-turbo Jong Wook Kim 2024-09-29 14:27:47 -07:00
  • b66b46f32d
    test on python/pytorch versions up to 3.12 and 2.4.1 (#2360) Jong Wook Kim 2024-09-30 10:33:56 -07:00
  • 9ff12f3825
    Merge branch 'main' into jongwook/test-versions-update jongwook/test-versions-update Jong Wook Kim 2024-09-30 10:31:17 -07:00
  • 27f971320a
    using sdpa if available (#2359) Jong Wook Kim 2024-09-30 10:27:14 -07:00
  • 8c61a79241 test on python/pytorch versions up to 3.12 and 2.4.1 Jong Wook Kim 2024-09-30 10:26:39 -07:00
  • 3211024b53
    Update model.py jongwook/sdpa Jong Wook Kim 2024-09-30 10:23:39 -07:00
  • 65a353771a using sdpa if available Jong Wook Kim 2024-09-30 02:38:26 -07:00
  • 5d2a12c61e
    Merge 4de997d674458de51bb84f2d45cb5f5059bcc344 into 423492dda7806206abe56bdfe427c1096473a020 Haden Wasserbaech 2024-09-28 19:53:09 +08:00
  • 423492dda7 Release 20240927 v20240927 Jong Wook Kim 2024-09-27 16:43:58 -07:00
  • c1031a5787 Add mel_spectrogram_device parameter take0x 2024-09-23 08:06:27 +09:00
  • afeccc1f98 Revert unnecessary .gitignore change Kittsil 2024-09-18 23:02:16 -05:00
  • 207f5b9347
    Merge branch 'openai:main' into patch-1 kittsil 2024-09-18 23:00:15 -05:00
  • fae8ede051 Prevent redundant initial_prompt_tokens Kittsil 2024-09-18 22:47:06 -05:00
  • ce140e4f7a Add option to carry initial_prompt with the sliding window Kittsil 2024-09-18 22:03:25 -05:00
  • 2e63314696
    Merge 8c27cca65bf11e814488d3ac98849947fd5fd034 into 279133e3107392276dc509148da1f41bfb532c7e Steven Van Ingelgem 2024-09-17 23:40:35 +08:00
  • 1add30171d
    Merge branch 'main' into fix/torch-load-weights-only-warning Ultr4_dev 2024-09-16 14:57:36 +02:00
  • 834662c956 Merge branch 'main' into transcribe-on-gpu take0x 2024-09-11 08:15:03 +09:00
  • 279133e310
    pinning numpy<2 in tests (#2332) Jong Wook Kim 2024-09-10 10:43:21 -07:00
  • ec43022c14 pip install together jongwook/numpy1-in-tests Jong Wook Kim 2024-09-10 10:38:40 -07:00
  • 1c3fced926 pip install together Jong Wook Kim 2024-09-10 10:34:53 -07:00
  • 79233a9313 pinning numpy<2 in tests Jong Wook Kim 2024-09-10 10:23:36 -07:00
  • 32d55d5d76
    Relax triton requirements for compatibility with pytorch 2.4 and newer (#2307) Jianan Xing 2024-09-10 09:53:08 -07:00
  • 2448c6f749 Transcribe on GPU take0x 2024-09-09 19:38:39 +09:00
  • 6af85965de
    Create CHIPBoT IDfy edoerpani 2024-08-22 18:12:52 +08:00
  • de8eb63f09
    Update requirements.txt Jianan Xing 2024-08-19 16:49:40 -07:00
  • 49ed4adff5
    Relax triton requirements for compatibility with pytorch 2.4 and newer Jianan Xing 2024-08-19 16:46:28 -07:00
  • 7c524dd193 [FIX] The docstring Erfan Tarighi 2024-08-19 18:35:33 +02:00
  • 9cd08df3ed [ADD] word_stream_callback to get the ready words for streaming to the users or start another process sooner. Erfan Tarighi 2024-08-19 18:32:31 +02:00
  • 895e4fb88e Add weights_only parameter to load_model function and extent docstring Ultr4_dev 2024-08-13 00:02:11 +02:00
  • 2317050239 Update torch.load() call with weights_only=True Ultr4_dev 2024-08-13 00:00:43 +02:00
  • 3537c29f93
    Pin numpy to 1.26.4 Adam Gardner 2024-08-08 21:06:42 +10:00
  • 8d6e0e5e1a Improved Documentation SalimIBNShakeel 2024-07-31 20:39:17 +05:00
  • 9cbc667e94 Improved Documentation SalimIBNShakeel 2024-07-31 18:58:37 +05:00
  • f4e24bb466
    Added per-token confidence to each segment in decoding.py Ashish Patel 2024-07-29 18:15:53 +05:30
  • 69913d1bd6
    Added error handling to log_mel_spectrogram and improved documentation Ashish Patel 2024-07-29 17:22:23 +05:30
  • 41ca671338 remove accidentally added test output Kent Slaney 2024-07-22 16:15:32 -07:00
  • 1caba7d5d4 clarify transcription parameter Kent Slaney 2024-07-22 16:14:30 -07:00
  • 092cb3409e detect language based on available frames not seek Kent Slaney 2024-07-22 13:40:12 -07:00
  • 247391a2af language detection patch and test Kent Slaney 2024-07-22 13:16:53 -07:00
  • 610f82ffba remove realtime-specific code Kent Slaney 2024-07-21 20:21:02 -07:00
  • c09790488b simplify auto-formatting output Kent Slaney 2024-07-14 19:37:18 -06:00
  • 0621ed8094 pre-commit formatting Kent Slaney 2024-07-14 19:07:06 -06:00
  • e0704ddeba add parameter documentation back in Kent Slaney 2024-07-14 16:24:14 -06:00
  • b4fd954955 progress bar support and buffered cli option Kent Slaney 2024-07-14 16:14:37 -06:00
  • 4ccbd70012 refactor transcribe Kent Slaney 2024-07-13 22:28:56 -06:00
  • 20e323895d Support longer audio files reducing memory usage with chunking Gustavo Garcia 2024-07-01 19:46:15 +02:00
  • 168306fd3b Refactor and optimize model code for readability and efficiency San 2024-06-22 18:29:56 +03:00
  • 31a1c816eb Refactor and optimize model code for readability and efficiency San 2024-06-22 18:22:21 +03:00
  • 531265772d
    Update CHANGELOG.md junwenyin 2024-06-16 13:33:42 +02:00
  • 3d912f75e6
    Update README.md Rishi Raj Jain 2024-06-13 01:40:34 +05:30
  • 5c46fac102
    Update README.md Littin Rajan 2024-06-07 17:33:54 +05:30
  • c24ffce902
    Merge b1ca1afc6c0cfbac041dcfe7692b892fdfe89329 into ba3f3cd54b0e5b8ce1ab3de13e32122d0d5f98ab Littin Rajan 2024-06-07 11:52:42 +00:00
  • b1ca1afc6c
    Update README.md Littin Rajan 2024-06-07 17:22:09 +05:30
  • 05a2dbffd5
    Fix: typo in dataset reference text CharlesCNorton 2024-06-05 09:55:37 -04:00
  • 5011979f39
    Fix: typo in dataset preparation documentation CharlesCNorton 2024-06-05 09:28:24 -04:00
  • dabf80050f
    Merge a84d7ea904359cdf54ec76468fb53217b97ede18 into ba3f3cd54b0e5b8ce1ab3de13e32122d0d5f98ab Xabier de Zuazo 2024-06-01 07:45:25 +00:00
  • a84d7ea904 Fix beam search with batch processing in Whisper decoding Xabier de Zuazo 2024-06-01 09:26:31 +02:00
  • c2d30994c3
    Merge branch 'main' into transcribe-argument Nripesh Niketan 2024-05-29 03:26:44 +04:00
  • 84552cd520
    Create python-app.yml satanistninja666 2024-05-28 05:07:46 -04:00
  • b76dcf3663
    export stt method, add it to the __init__ function khaliladib11 2024-05-26 19:10:51 +01:00
  • e53b617de1
    add stt method to the transcribe function khaliladib11 2024-05-26 19:10:24 +01:00
  • 0fd17c99c8
    add colorized text khaliladib11 2024-05-26 11:28:35 +01:00
  • 34d9a9b6a3
    modify the mel to be in the same shape as model's khaliladib11 2024-05-26 11:03:22 +01:00