Upload cleaned dataset with actual audio: 3126 transcribable items

README.md

@@ -0,0 +1,43 @@
+---
+dataset_info:
+  features:
+  - name: audio
+    dtype:
+      audio:
+        sampling_rate: 16000
+  - name: file_name
+    dtype: string
+  - name: folder
+    dtype: string
+  - name: transcription
+    dtype: string
+  - name: confidence
+    dtype: string
+  - name: trim_start_seconds
+    dtype: float32
+  - name: trim_end_seconds
+    dtype: float32
+  - name: reason_start
+    dtype: string
+  - name: reason_end
+    dtype: string
+  - name: has_incomplete_words
+    dtype: bool
+  - name: notes
+    dtype: string
+  - name: original_path
+    dtype: string
+  - name: batch_index
+    dtype: int32
+  splits:
+  - name: train
+    num_bytes: 903820704.782
+    num_examples: 3126
+  download_size: 854908727
+  dataset_size: 903820704.782
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+---

data/train-00000-of-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a52bc656fe4d6cd0155c3b10d79861bc874a2642aa43062ef25946f5eed1d85d
+size 423015876

data/train-00001-of-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0dddc372b3bf27fb7a251372e150a4de09aebb4dad7a5be963fd8221ec80524
+size 431892851