Upload dataset

README.md

@@ -4,10 +4,11 @@ dataset_info:
 - config_name: france
   features:
   - name: audio
-    dtype:
-      audio:
-        sampling_rate: 16000
-        decode: false
+    struct:
+    - name: bytes
+      dtype: binary
+    - name: path
+      dtype: 'null'
   - name: key
     dtype: string
   - name: country
@@ -19,29 +20,29 @@ dataset_info:
   - name: transcript_id
     dtype: string
   - name: start_seconds
-    dtype: float32
+    dtype: float64
   - name: end_seconds
-    dtype: float32
+    dtype: float64
   - name: duration_seconds
-    dtype: float32
+    dtype: float64
   - name: asr_transcript
     dtype: string
   - name: human_transcript
     dtype: string
   - name: cer
-    dtype: float32
+    dtype: float64
   - name: wer
-    dtype: float32
+    dtype: float64
   - name: original_transcript_start_idx
-    dtype: int32
+    dtype: int64
   - name: original_transcript_end_idx
-    dtype: int32
+    dtype: int64
   splits:
   - name: train
-    num_bytes: 189799373
-    num_examples: 393
-  download_size: 188673066
-  dataset_size: 189799373
+    num_bytes: 253185768910
+    num_examples: 530671
+  download_size: 251415142808
+  dataset_size: 253185768910
 - config_name: germany
   features:
   - name: audio

france/train-00000-of-00001.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:be0826edaea973a7b34010442b16ce0b477e9af8ab5a99948ef98507d9349344
-size 188673066