Fixed speaker and condition_on_prev

30s/test-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8a419f58149fd0f9f330d3874be1d08a4d618e664ae71410df58b6d1b0e1892f
-size 104733846
+oid sha256:3623046a97ee33d69cf35b32e57da90e51f0199e2da152df2dfc0be89aff4df8
+size 105200894

30s/train-00000-of-00002.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fbab0cf21cc59f9f06d89f45bd65dd9a7e1149a4393a710545576a466b3dbfa7
-size 506673760
+oid sha256:085d086422efc57578e77008a77248947d490f5dbfc7aa581355e26669b739f8
+size 505723004

30s/train-00001-of-00002.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:089bacf8cf6b572a952a7f677379e4c81a8c76ca6e696c605a63b74a4816c6df
-size 506694259
+oid sha256:9462ae37883513eb3377b3f7d369c15bbbb85dd4b6fd336fe1140ca9f224f27e
+size 506341012

30s/validation-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:33642e67df5b95a7cdd6f467972a7d44e60f62b90aae1ce8d60b9911249c890a
-size 105777914
+oid sha256:35c62cf88a499d2133f5f8ccb94ff514c4ed5dce4c9b03397f58d5508734f3e2
+size 105566982

30s_augment/train-00000-of-00002.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:94def5fa4ae0d5268f0eb3b55853153021eb99e611aab01dd3a7cff3c3e43392
-size 508144100

30s_augment/train-00001-of-00002.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:994d05eb3c8473246146f05a8de0b3df0f3229caecc6762b0f16831c65fc4f66
-size 506643252

30s_demucs/test-00000-of-00001.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e7a3a8cd0ded05a0833d58d946cd290379dab9929e835718895e0b5ecabf5973
-size 111773726

30s_demucs/train-00000-of-00002.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:69d92b10a3c46d0942af4c2f65db968f4cb7c088cb11c91cc88932b6eb2391cf
-size 541349063

30s_demucs/train-00001-of-00002.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7c9aafa33523b2c55e1526321cdfc17a6bc3cfb1af673f16f36847bd313afb6f
-size 538618160

30s_demucs/validation-00000-of-00001.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e78d86d51ee497881a0853498c33051db37e42e8dba38ebd7731f44740eac2a2
-size 113051374

README.md

@@ -6,74 +6,6 @@ license: cc-by-sa-4.0
 size_categories:
 - 10K<n<100K
 dataset_info:
-- config_name: 30s
-  features:
-  - name: id
-    dtype: string
-  - name: audio
-    dtype:
-      audio:
-        sampling_rate: 16000
-  - name: transcription
-    dtype: string
-  - name: condition_on_prev
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 1131899415.68
-    num_examples: 1315
-  - name: validation
-    num_bytes: 118639586.0
-    num_examples: 139
-  - name: test
-    num_bytes: 120445541.0
-    num_examples: 140
-  download_size: 1223879779
-  dataset_size: 1370984542.68
-- config_name: 30s_augment
-  features:
-  - name: id
-    dtype: string
-  - name: audio
-    dtype:
-      audio:
-        sampling_rate: 16000
-  - name: transcription
-    dtype: string
-  - name: duration
-    dtype: float32
-  - name: condition_on_prev
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 1139055524.508
-    num_examples: 1332
-  download_size: 1014787352
-  dataset_size: 1139055524.508
-- config_name: 30s_demucs
-  features:
-  - name: id
-    dtype: string
-  - name: audio
-    dtype:
-      audio:
-        sampling_rate: 16000
-  - name: transcription
-    dtype: string
-  - name: condition_on_prev
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 1131902020.68
-    num_examples: 1315
-  - name: validation
-    num_bytes: 118639860.0
-    num_examples: 139
-  - name: test
-    num_bytes: 120445813.0
-    num_examples: 140
-  download_size: 1304792323
-  dataset_size: 1370987693.68
 - config_name: default
   features:
   - name: id
@@ -109,34 +41,22 @@ dataset_info:
   download_size: 1223500329
   dataset_size: 1227500846.4050002
 configs:
-- config_name: 30s
-  data_files:
-  - split: train
-    path: 30s/train-*
-  - split: validation
-    path: 30s/validation-*
-  - split: test
-    path: 30s/test-*
-- config_name: 30s_augment
-  data_files:
-  - split: train
-    path: 30s_augment/train-*
-- config_name: 30s_demucs
-  data_files:
-  - split: train
-    path: 30s_demucs/train-*
-  - split: validation
-    path: 30s_demucs/validation-*
-  - split: test
-    path: 30s_demucs/test-*
 - config_name: default
   data_files:
-  - split: train
-    path: data/train-*
-  - split: test
-    path: data/test-*
-  - split: validation
-    path: data/validation-*
+  - path: data/train-*
+    split: train
+  - path: data/test-*
+    split: test
+  - path: data/validation-*
+    split: validation
+- config_name: 30s
+  data_files:
+  - path: 30s/train-*
+    split: train
+  - path: 30s/validation-*
+    split: validation
+  - path: 30s/test-*
+    split: test
 ---
 # Dataset Card for Dataset Name
 
@@ -145,6 +65,18 @@ This dataset is derived from CAiRE/ASCEND. More information is available at http
 - Removed 嗯 呃 um uh
 - Resolved [UNK]'s using whisper-medium
 
+## Usage
+- Default utterances with cleaned transcripts
+```python
+from datasets import load_dataset
+data = load_dataset("georgechang8/ASCEND_CLEAN")  # add split="train" for train set, etc.
+```
+- Concatenated 30s utterances with cleaned transcripts
+  - https://github.com/George0828Zhang/distil-whisper/blob/main/training/run_concatenate.py
+```python
+data = load_dataset("georgechang8/ASCEND_CLEAN", "30s")  # add split="train" for train set, etc.
+```
+
 ## Dataset Details
 
 ### Dataset Description
@@ -287,8 +219,6 @@ def fix_unk(sample, adjusted_dict):
         adjustment = transcription
     return {"transcription": adjustment}
 
-
-
 data = DatasetDict({
     split: data_raw[split].map(lambda x: fix_unk(x, adjusted_transcripts[split]), load_from_cache_file=False)
     for split in data_raw