Add SpokenQAforIC dataset for MTEB audio classification evaluation

README.md

@@ -0,0 +1,25 @@
+---
+dataset_info:
+  features:
+  - name: audio
+    dtype:
+      audio:
+        sampling_rate: 16000
+  - name: label
+    dtype: string
+  - name: instruction
+    dtype: string
+  - name: file
+    dtype: string
+  splits:
+  - name: test
+    num_bytes: 416693860.27
+    num_examples: 6121
+  download_size: 406982504
+  dataset_size: 416693860.27
+configs:
+- config_name: default
+  data_files:
+  - split: test
+    path: data/test-*
+---

data/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:09f444d6591d8ee8852783693ee881833a6d7df231e88f5679b1c15ee4cb2606
+size 406982504