Upload 2 files
Browse filesNewly generated dev and train datasets with more focus on the 40-60 percent overlapped speech.
data/dev_no_aug_convo_focus/1_or_2_mix_extract_no_augmentations.tar
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec7ce838e41f58bf6c6440fbe552570f353f070453c5461c40b6c02806874b2b
|
| 3 |
+
size 3799889920
|
data/train_no_aug_convo_focus/1_or_2_mix_extract_no_augmentations.tar
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a9316a8c85f1002086b0e84feeef9f6f58becd6c4427c6cb39828d765c45b788
|
| 3 |
+
size 15391836160
|