Commits

Breadcrumbs

History for

onmain

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 24, 2023

• Updated ErrorMessage.jsx

  Show description for c6bc037
  ethanzrd

  committed
  c6bc037

  Copy full SHA for c6bc037

• Updated delay calculation

  Show description for e082b1b
  ethanzrd

  committed
  e082b1b

  Copy full SHA for e082b1b

Commits on Aug 15, 2023

• Enhanced codebase clarity and adhered to best practices by significantly reducing the reliance on plain strings. Implemented data type indicators in functions handling audio data, incorporated asse…

  Show description for c825d5c
  ethanzrd

  committed
  c825d5c

  Copy full SHA for c825d5c

• Enhanced audio pre-processing by implementing voice activity detection for selecting speech-containing audio chunks. Also, addressed end-of-stream behavior for sequential mode and cleanup behavior …

  Show description for 33534eb
  ethanzrd

  committed
  33534eb

  Copy full SHA for 33534eb

Commits on Aug 13, 2023

• Implemented error handling for transcription call to resolve silence suppression issue. Resolved the 'Unknown Speaker' label incorrectly appearing in the client.

  ethanzrd

  committed
  eb1f200

  Copy full SHA for eb1f200

Commits on Aug 11, 2023

• Fix incorrect import statements

  ethanzrd

  committed
  b782aee

  Copy full SHA for b782aee

Commits on Aug 10, 2023

• Revised installation script to generate a fresh Conda environment named "whisper-playground," integrated configuration options for the transcription device and compute type, improved clarity of REA…

  Show description for 7681c13
  ethanzrd

  committed
  7681c13

  Copy full SHA for 7681c13

Commits on Aug 8, 2023

• Implemented cleanup process for sequential transcription. Resolved some issues preventing local application execution, and optimized the installation script.

  ethanzrd

  committed
  86cfcad

  Copy full SHA for 86cfcad

Commits on Aug 5, 2023

• Refined speaker identification with error handling in sequential transcription

  Show description for cc01bef
  ethanzrd

  committed
  cc01bef

  Copy full SHA for cc01bef

Commits on Aug 2, 2023

• Remove redundant line and update README.md

  ethanzrd

  committed
  ac079b9

  Copy full SHA for ac079b9

• Implemented the ability to transcribe audio sequentially, allowing the application to automatically re-transcribe the entire content every x seconds (transcription timeout). Utilized pyannote's aud…

  Show description for 5a68efb
  ethanzrd

  committed
  5a68efb

  Copy full SHA for 5a68efb

Commits on Jul 31, 2023

• Optimized model initialization process by choosing English-specific models when possible.

  ethanzrd

  committed
  8d0f62b

  Copy full SHA for 8d0f62b

• Refined transcription process with faster-whisper + stable-ts combo, reducing memory usage, eliminating timestamp hallucination bug, and accelerating transcription speed. Minimized failed speaker i…

  Show description for f2fbadd
  ethanzrd

  committed
  f2fbadd

  Copy full SHA for f2fbadd

Commits on Jul 27, 2023

• Enhanced responsiveness and graceful shutdown for Whisper models. This commit offloads blocking initialization of Whisper models to a separate thread, as well as disconnection and stream ending ope…

  Show description for 47ee43d
  ethanzrd

  committed
  47ee43d

  Copy full SHA for 47ee43d

• Refactor: Improve Clarity, Readability, and Error Handling

  Show description for e601f9d
  ethanzrd

  committed
  e601f9d

  Copy full SHA for e601f9d

Commits on Jul 26, 2023

• Fixed installation script not working and added error handling when removing clients from the server.

  ethanzrd

  committed
  7c815e7

  Copy full SHA for 7c815e7

• Implemented real-time audio transcription with speaker diarization using sockets and Diart library.

  ethanzrd

  committed
  3292c77

  Copy full SHA for 3292c77

Commits on Nov 2, 2022

• Improved verbose formatting output, added timing

  chrismbirmingham

  committed
  ce1e7fe

  Copy full SHA for ce1e7fe

• bugfix- predicted_text in verbose case

  chrismbirmingham

  committed
  fe17e5d

  Copy full SHA for fe17e5d

Commits on Oct 5, 2022

• Language and model size can now be configured via the frontend

  saharmor

  committed
  efbab58

  Copy full SHA for efbab58

Commits on Oct 4, 2022

• fullstack playground with transcription timeout

  saharmor

  committed
  3d56a20

  Copy full SHA for 3d56a20

• Whisper standalone + webservice

  saharmor

  committed
  be8185f

  Copy full SHA for be8185f

Loading