Clean debug strings.

VideoLLaMA2/videollama2/model/videollama2_arch.py

@@ -117,7 +117,6 @@ class Videollama2MetaForCausalLM(ABC):
 
         data_batch = []
         for i, (data, modal) in enumerate(images):
-            print(data.shape, modal)
             if modal == 'image':
                 data = data.expand(num_frames, -1, -1, -1)
             else:
@@ -126,8 +125,6 @@ class Videollama2MetaForCausalLM(ABC):
 
         data_batch = torch.stack(data_batch, dim=0)
 
-        print(data_batch.shape)
-
         assert len(data_batch.size()) == 5
         batch_size = data_batch.size(0)