Initialize EchoDialogue dataset card
Browse files
README.md
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
configs:
|
| 3 |
+
- config_name: single_turn
|
| 4 |
+
data_files:
|
| 5 |
+
- split: train
|
| 6 |
+
path: single_turn/train/*.tar
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
# EchoDialogue
|
| 10 |
+
|
| 11 |
+
EchoDialogue is an emotional spoken-dialogue dataset with single-turn and
|
| 12 |
+
multi-turn configurations. The single-turn split is being prepared as
|
| 13 |
+
anonymous WebDataset shards.
|
| 14 |
+
|
| 15 |
+
Each single-turn sample contains query audio, response audio, dialogue text,
|
| 16 |
+
emotion annotations, speaking-style annotations, and reasoning metadata.
|
| 17 |
+
Original UUIDs, storage paths, and source filenames are not published.
|
| 18 |
+
|
| 19 |
+
The multi-turn configuration will be added separately.
|