Upload train_vb_29nqe0uh_epoch=115.ckpt

For the speech enhancement task, we offer pretrained checkpoints for models that have been trained on the VoiceBank-DEMAND and WSJ0-CHiME3 datasets, as described in our journal paper [2].

SGMSE+ trained on VoiceBank-DEMAND: gdown 1_H3EXvhcYBhOZ9QNUcD5VZHc6ktrRbwQ
SGMSE+ trained on WSJ0-CHiME3: gdown 16K4DUdpmLhDNC7pJhBBc08pkSIn_yMPi

pretrained_checkpoints/speech_enhancement/train_vb_29nqe0uh_epoch=115.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3875747b5646092d5c556bae68e5af639e2c1f45f009c669f379cd4d415cbd8
+size 1312806643