Hi, I noticed that you updated your training code last week. I noticed that in the new code, the mode selection is missing the "audiovisual" option, and I wonder how I can continue training the AVSR task? If the "video" mode is applicable, does it perform AVSR tasks correctly?
Hi, I noticed that you updated your training code last week. I noticed that in the new code, the mode selection is missing the "audiovisual" option, and I wonder how I can continue training the AVSR task? If the "video" mode is applicable, does it perform AVSR tasks correctly?