RuntimeError: torch.cat(): expected a non-empty list of Tensors

[herbiel]

Number of Classes is 3
Now load features from /data/sls/scratch/yuangong/whisper-a/feat_as_full/whisper_tiny
Dataset has 31 samples
Using Label Smoothing: 0.0
Using Following Mask: 0 Freq, 0 Time
Using Mix-up with Rate 0.000000
Now Process as-full
Number of Classes is 3
Now load features from /data/sls/scratch/yuangong/whisper-a/feat_as_eval/whisper_tiny
val_loader configuration!!!!!!!! /data/ph/test.json
test in tin!!!!!!!!!!!!!1y
lw_tr_1_8 tiny 4 384

Creating experiment directory: ./exp/test-as-full-whisper-whisper-high-lw_tr_1_8-tiny-5e-5-15-0.75-bs48-ldaFalse-mix0.5-0-10
Now starting training for 30 epochs
running on cuda
Total parameter number is : 3.550 million
Total trainable parameter number is : 3.550 million
The learning rate scheduler starts at 15 epoch with decay rate of 0.750 every 5 epoches
now training with as-full, main metrics: mAP, loss function: BCEWithLogitsLoss(), learning rate scheduler: <torch.optim.lr_scheduler.MultiStepLR object at 0x7f400ed87490>
current #steps=0, #epochs=1
start training...

2024-01-17 08:59:12.731168
current #epochs=1, #steps=0
test shape
torch.Size([48, 4, 25, 384])
start validation
111111111111
11112222211
val_loader is <torch.utils.data.dataloader.DataLoader object at 0x7f400ed878e0>
[] []
Traceback (most recent call last):
File "/opt/whisper/whisper-at/src/whisper_at_train/./run.py", line 162, in
train(audio_model, train_loader, val_loader, args)
File "/opt/whisper/whisper-at/src/whisper_at_train/traintest.py", line 137, in train
stats, valid_loss = validate(audio_model, test_loader, args)
File "/opt/whisper/whisper-at/src/whisper_at_train/traintest.py", line 241, in validate
audio_output = torch.cat(A_predictions)
RuntimeError: torch.cat(): expected a non-empty list of Tensors

what's problem with this train stage