Add audio denoising dataset with original, noisy, and denoised pairs

README.md

@@ -0,0 +1,33 @@
+---
+dataset_info:
+  features:
+  - name: sample_id
+    dtype: string
+  - name: original_audio
+    dtype:
+      audio:
+        sampling_rate: 16000
+  - name: complex_noisy
+    dtype:
+      audio:
+        sampling_rate: 16000
+  - name: white_noisy
+    dtype:
+      audio:
+        sampling_rate: 16000
+  - name: denoised
+    dtype:
+      audio:
+        sampling_rate: 16000
+  splits:
+  - name: train
+    num_bytes: 14438266.0
+    num_examples: 10
+  download_size: 13999754
+  dataset_size: 14438266.0
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+---

data/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90bb4d7dec35d74547a47f36fe8d4cab4c49639a5cb06d9a94d1452613931b56
+size 13999754