10 days ago
-
v3-lite weights, evaluation and training log
-
Add config.json (download counting)
11 days ago
-
Card: gertelrina namespace, corrected music false-positive numbers; eval: music_false_positives.json
-
Model card wording after review; eval: cross-evaluation, NISQA runs, speed, FSD50K noise-pool list
12 days ago
-
Model card: v3.1 changes as bullet points
-
Model card: thresholds of the false-alarm columns
-
Model card: time resolution wording
-
Model card: speed and benchmark hardware
Rina G
13 days ago
-
Model card: install line
Rina G
-
Demo and report screenshot from v3.1
Rina G
-
v3.1: fine-tuned with FSD50K noise and babble (fewer music false alarms); eval reports
Rina G
14 days ago
-
Demo GIF re-recorded
Rina G
-
Demo: recording of the report file view
Rina G
-
Model card: GitHub link and badges
Rina G
-
Add v1 and v2 weights, evaluation reports in eval/
Rina G
-
AADD v3: Whisper-small frame-level degradation detector (12 classes, 20 ms)
Rina G
-
initial commit
Rina G