Update README.md

README.md

@@ -15,14 +15,14 @@ configs:
   - split: aru
     path: data/aru/utterances*.jsonl
   - split: cmu_arctic
-    path: data/cmu-arctic/utterances*.jsonl
+    path: data/cmu_arctic/utterances*.jsonl
 - config_name: utterance_embeddings
   version: 1.0.0
   data_files:
   - split: aru
     path: data/aru/utterance_embeddings*.jsonl
   - split: cmu_arctic
-    path: data/cmu-arctic/utterance_embeddings*.jsonl
+    path: data/cmu_arctic/utterance_embeddings*.jsonl
 - config_name: speaker_embeddings
   version: 1.0.0
   data_files: data/*/speaker_embeddings*.jsonl
@@ -63,7 +63,9 @@ dataset_info:
     dtype: string
   - name: name
     dtype: string
-  - name: speaker
+  - name: duration
+    dtype: float32
+  - name: speaker_id
     dtype: string
   - name: dataset_id
     dtype: string

data/aru/{speaker_embeddings.jsonl → speaker_embeddings_001.jsonl}

@@ -1,2 +1 @@
 { "speaker_id": "001", "dataset_id": "001", "model_id": "001", "embedding": [1.0, 2.0, 3.0] }
-{ "speaker_id": "001", "dataset_id": "001", "model_id": "001", "embedding": [1.0, 2.0, 3.0] }

data/aru/speaker_embeddings_002.jsonl

@@ -0,0 +1 @@
+{ "speaker_id": "001", "dataset_id": "001", "model_id": "002", "embedding": [1.0, 2.0, 3.0] }

data/aru/utterances.jsonl

@@ -1,2 +1,2 @@
-{ "id": "001", "name": "aru-file1.wav", "speaker": "001", "dataset_id": "001" }
-{ "id": "002", "name": "aru-file2.wav", "speaker": "001", "dataset_id": "001" }
+{ "id": "001", "name": "aru-file1.wav", "duration": 3.25, "speaker_id": "001", "dataset_id": "001" }
+{ "id": "002", "name": "aru-file2.wav", "duration": 1.53, "speaker_id": "001", "dataset_id": "001" }

data/cmu-arctic/utterances.jsonl

@@ -1,2 +0,0 @@
-{ "id": "001", "name": "cmu-arctic-file1.wav", "speaker": "002", "dataset_id": "002" }
-{ "id": "002", "name": "cmu-arctic-file2.wav", "speaker": "002", "dataset_id": "002" }

data/{cmu-arctic/speaker_embeddings.jsonl → cmu_arctic/speaker_embeddings_001.jsonl}

@@ -1,2 +1 @@
 { "speaker_id": "002", "dataset_id": "002", "model_id": "001", "embedding": [1.0, 2.0, 3.0] }
-{ "speaker_id": "002", "dataset_id": "002", "model_id": "001", "embedding": [1.0, 2.0, 3.0] }

data/cmu_arctic/speaker_embeddings_002.jsonl

@@ -0,0 +1 @@
+{ "speaker_id": "002", "dataset_id": "002", "model_id": "002", "embedding": [1.0, 2.0, 3.0] }

data/cmu_arctic/utterances.jsonl

@@ -0,0 +1,2 @@
+{ "id": "001", "name": "cmu-arctic-file1.wav", "duration": 2.3, "speaker_id": "002", "dataset_id": "002" }
+{ "id": "002", "name": "cmu-arctic-file2.wav", "duration": 1.53, "speaker_id": "002", "dataset_id": "002" }