claudetools

azcomputerguru/claudetools

Fork 0

Commit Graph

Author	SHA1	Message	Date
Mike Swanson	e9ac607500	radio show: co-host voice profile, Q&A extraction fixes, archive index - Build Tom (co-host) voice profile (44 embeddings, 0.698 similarity to Mike) - diarizer.py: add CO-HOST speaker label for cohost-role profiles - voice_profiler.py: emit "Cohost: <name>" label for cohost role - qa_extractor.py: overlap resolution at load time (midpoint boundary split), 4s CALLER-preference threshold, turn-based caller-intro lookback (2 HOST turns), _preceded_by_caller_intro() helper, _PHONE_GREETING pattern, 751-1041 + "we'll get your problem solved" promo signatures - benchmark.py: use src.transcriber.transcribe with batch_size=16 - add index_test_episodes.py and build_cohost_profile.py scripts - add .gitignore (exclude episodes, transcripts, *.db, .venv) - session log: 2026-04-27-qa-extraction-cohost-indexing.md Result: 2016-s8e43 drops from 12 false-positive Q&A pairs to 2 real caller pairs. archive.db: 6 episodes, 762 segments, 10 Q&A pairs, FTS5 search verified. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-04-27 14:41:04 -07:00
Mike Swanson	6cc9043b8e	Audio processor: validated voice profiling accuracy, tuned threshold - Fine-grained speaker analysis (3s windows, 1s hop) across 42min episode - Host voice: 0.90-0.98 similarity (clear positive match) - Callers: 0.65-0.68 (correctly below threshold) - Produced audio/clips: 0.53-0.65 (correctly identified as non-host) - Co-host/other speakers: 0.56-0.62 (correctly identified) - Tuned host_match_threshold from 0.75 to 0.83 based on empirical data - Cross-referenced dips with transcript: correctly identifies callers, show intros, played audio clips, and station breaks - Batch transcription of 7 additional training episodes in progress Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-21 12:48:25 -07:00

Author

SHA1

Message

Date

Mike Swanson

e9ac607500

radio show: co-host voice profile, Q&A extraction fixes, archive index

- Build Tom (co-host) voice profile (44 embeddings, 0.698 similarity to Mike)
- diarizer.py: add CO-HOST speaker label for cohost-role profiles
- voice_profiler.py: emit "Cohost: <name>" label for cohost role
- qa_extractor.py: overlap resolution at load time (midpoint boundary split),
  4s CALLER-preference threshold, turn-based caller-intro lookback (2 HOST turns),
  _preceded_by_caller_intro() helper, _PHONE_GREETING pattern,
  751-1041 + "we'll get your problem solved" promo signatures
- benchmark.py: use src.transcriber.transcribe with batch_size=16
- add index_test_episodes.py and build_cohost_profile.py scripts
- add .gitignore (exclude episodes, transcripts, *.db, .venv)
- session log: 2026-04-27-qa-extraction-cohost-indexing.md

Result: 2016-s8e43 drops from 12 false-positive Q&A pairs to 2 real caller pairs.
archive.db: 6 episodes, 762 segments, 10 Q&A pairs, FTS5 search verified.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

2026-04-27 14:41:04 -07:00

Mike Swanson

6cc9043b8e

Audio processor: validated voice profiling accuracy, tuned threshold

- Fine-grained speaker analysis (3s windows, 1s hop) across 42min episode
- Host voice: 0.90-0.98 similarity (clear positive match)
- Callers: 0.65-0.68 (correctly below threshold)
- Produced audio/clips: 0.53-0.65 (correctly identified as non-host)
- Co-host/other speakers: 0.56-0.62 (correctly identified)
- Tuned host_match_threshold from 0.75 to 0.83 based on empirical data
- Cross-referenced dips with transcript: correctly identifies callers,
  show intros, played audio clips, and station breaks
- Batch transcription of 7 additional training episodes in progress

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-03-21 12:48:25 -07:00

2 Commits