Advanced Search
Search Results
41 total results found
From Recording to Summary
# Einmalig: Prompt-Datei aus YAML generieren uv run --with pyyaml python3 bin/build-whisper-prompt.py --game packs_and_prejudice # → meta/packs_and_prejudice.prompt editieren, Ballast löschen # Danach wie gewohnt: transcribe-audio.sh --en --game packs_and_p...
build-whisper-prompt.py
#!/usr/bin/env python3 """ build-whisper-prompt.py Extracts primary keys from a TranscriptOMatic YAML meta file and writes them one per line to a .prompt file for use as a whisper vocabulary hint. The output is a starting point — edit it manually to remove co...
Building Consistency
Testing several phonetic spellings for names with a lot of errors Maybe testing several spellings for a name Using not the 'most correct' version, but the most useful one Using a section of the game's YAML document to note down which variant worked with what ...
summarize-meeting.sh
#!/usr/bin/env bash set -euo pipefail # ------------------------------------------------------------ # summarize-meeting.sh # Post-session summarization on Mac using ollama # Usage: summarize-meeting.sh [--de|--en] [--game <slug>] [<transcript.txt>] # ...
build-summary-context.py
#!/usr/bin/env python3 """ build-summary-context.py Extracts summarizer-relevant fields from a TranscriptOMatic YAML meta file, stripping transcription-specific fields (aliases, safe, lang, deedname, etc.) that confuse LLMs when used as summary context. Outpu...