Conversation
…wson/PDR_AI_v2 into feature/speech-to-text
… for viewing audio, and also updated the transcription file. transcription now shows the audio playing with timestamps and allows the user to click on a time stamp or line of audio and have the audio clip jump to that spot.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
|
@codex review |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: a8d29a8ec7
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| volumes: | ||
| postgres_data: | ||
| seaweedfs_data: | ||
| sidecar_models: |
There was a problem hiding this comment.
In docker-compose.yml, the seaweedfs service still mounts seaweedfs_data:/data (line 143), but the top-level volumes block no longer declares seaweedfs_data. When running with --profile local-storage, Compose treats this as an undeclared named volume and fails to bring up that profile, so local S3 storage is broken until the volume declaration is restored.
Useful? React with 👍 / 👎.
| /** Infer display type from document for viewer rendering */ | ||
| export function getDocumentDisplayType(doc: { url: string; title: string; mimeType?: string }): DocumentDisplayType { | ||
| // Check title first — transcription documents are stored as text/plain but should render as audio | ||
| if (doc.title.toLowerCase().includes("(transcription)")) return "audio"; |
There was a problem hiding this comment.
This unconditional title check routes any document whose title contains (transcription) to the audio viewer before MIME/extension checks run. That misclassifies non-audio files (for example, PDFs or text docs with that phrase in the title), causing DocumentViewer to render AudioViewer and breaking normal preview behavior for those documents.
Useful? React with 👍 / 👎.
| ports: | ||
| - "8000:8000" | ||
| healthcheck: | ||
| test: ["CMD", "curl", "-f", "http://localhost:8000/health"] |
There was a problem hiding this comment.
The new healthcheck executes curl -f http://localhost:8000/health, but the sidecar image (see sidecar/Dockerfile) installs build-essential and ffmpeg only, so curl is not provided by this Dockerfile. In that setup the healthcheck returns command-not-found and marks the container unhealthy even when the API is actually running.
Useful? React with 👍 / 👎.
3 participants
Audio transcription using OpenAI whisper on python sidecar. Provides transcriptions with interactive timestamps of MP3 and MP4