status update

README.md

@@ -14,19 +14,19 @@ Environment-aware TTS training corpus, built from 4 speaker-diarization sources:
 M3SD (770 h) · AISHELL-4 (120 h) · MSDWILD (80 h) · CHiME-6 (40 h).
 Each row: `(environment_audio_source, speaker_audio_source, text, speech)` as 16 kHz mono FLAC bytes, with provenance.
 
-**Last status push:** 2026-05-14T18:18:24+00:00Z · status_updater since 2026-05-14T06:25:15+00:00Z
+**Last status push:** 2026-05-14T18:23:38+00:00Z · status_updater since 2026-05-14T06:25:15+00:00Z
 
 ## ⏳ ETA to completion
 
 | | wall-clock estimate |
 | --- | --- |
-| **Total remaining** | **13.1 h** |
+| **Total remaining** | **13.2 h** |
 | Current stage (MSDWILD) | 0 min |
 | AISHELL-4 + CHiME-6 stages | 2.3 h |
 | M3SD baidu download | — |
 | M3SD processing | 10.9 h |
 
-_Throughput so far: **71.0 audio-hours per wall-hour** (954.4 h processed in 11.9 h elapsed)_
+_Throughput so far: **70.6 audio-hours per wall-hour** (955.7 h processed in 12.0 h elapsed)_
 
 ## Live pipeline status
 
@@ -35,7 +35,7 @@ _Throughput so far: **71.0 audio-hours per wall-hour** (954.4 h processed in 11.
 | pipeline alive | ✅ |
 | GPU utilisation | 0% |
 | GPU VRAM used | 18707 / 97 887 MiB |
-| disk used | 189 / 200 GiB |
+| disk used | 191 / 200 GiB |
 | current stage | CHIME6 |
 
 ## Dataset progress
@@ -44,12 +44,12 @@ _Throughput so far: **71.0 audio-hours per wall-hour** (954.4 h processed in 11.
 | --- | ---: | ---: | --- |
 | MSDWILD | 3113 | 28432 | ✅ done |
 | AISHELL-4 | 145 | 35252 | ✅ done |
-| CHiME-6 | 17 | 13010 | live |
+| CHiME-6 | 18 | 13010 | live |
 | M3SD | 1372 | 211809 | ✅ done |
 
 ## Throughput
 
-- Records / min: **5.19**
+- Records / min: **5.15**
 - ETA for MSDWILD stage: **~5 min**
 - M3SD baidu download: **68.6 GB / 68.63 GB**