Add quality confidence columns (26169 clean, 214 flagged, avg confidence 0.948)

README.md

@@ -15,6 +15,59 @@ multilinguality:
 - highly_multilingual
 size_categories:
 - 10K<n<100K
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+dataset_info:
+  features:
+  - name: audio
+    dtype: audio
+  - name: video_id
+    dtype: string
+  - name: sentence
+    dtype: string
+  - name: segments_en
+    dtype: string
+  - name: segments_ms
+    dtype: string
+  - name: total_speakers
+    dtype: int64
+  - name: sentence_en
+    dtype: string
+  - name: sentence_ms
+    dtype: string
+  - name: sentence_zh_hans
+    dtype: string
+  - name: sentence_zh_hant
+    dtype: string
+  - name: sentence_ru
+    dtype: string
+  - name: sentence_id
+    dtype: string
+  - name: sentence_ar
+    dtype: string
+  - name: sentence_ja
+    dtype: string
+  - name: sentence_ko
+    dtype: string
+  - name: leakage_avg
+    dtype: float64
+  - name: leakage_max
+    dtype: float64
+  - name: overlap_pct
+    dtype: float64
+  - name: single_speaker_pct
+    dtype: float64
+  - name: quality_confidence
+    dtype: float64
+  splits:
+  - name: train
+    num_bytes: 69356297972
+    num_examples: 26383
+  download_size: 69327852389
+  dataset_size: 69356297972
 ---
 
 # khursanirevo/multiturn_ks