search

WeidiXie / VGG-Speaker-Recognition

Utterance-level Aggregation For Speaker Recognition In The Wild
362 stars 98 forks source link

Error while attempting to train the model #67

Open aserzhqa54zwsz opened 3 years ago

aserzhqa54zwsz commented 3 years ago

Hello, I am trying to make this implementation work, but I am having trouble. I did the following (running Ubuntu 20 if that helps):

  1. I cloned the repository.
  2. Installed adequate module versions as given in req.txt (I'll list them if needed).

Now I am attempting to train the model over voxCeleb2, but I am experiencing trouble with the meta files: IOError: [Errno 2] No such file or directory: '../meta/voxlb2_val.txt'

I am not sure what this file is supposed to be, but appears to be missing from the meta folder. Any help would greatly be appreciated.

Fan0fan commented 2 years ago

I have met the same problem. Have you ever solve it? Please tell me.

bameroncaird commented 2 years ago

Did either of you solve this issue? I see from the closed issues that we should manually generate the validation data from '../meta/voxlb2_train.txt', but it would save me some time if I could find the file elsewhere.

bameroncaird commented 2 years ago

Solved: https://github.com/hmen97/VGG-Vox/tree/master/meta