ASiT input shape

[alexanderwerning]

Hi,
is there a specific reason why the input shape seems to be [time, mel_bins] in vision_transformer.py:

PretrainedSED/models/asit/vision_transformer.py

Lines 118 to 123 in 1aa47e4

	class VisionTransformer(nn.Module):
	""" Vision Transformer """
	def __init__(self, audio_size=[1024, 128], patch_size=[16, 16], in_chans=3, num_classes=0, embed_dim=768, depth=12,
	num_heads=12, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop_rate=0., attn_drop_rate=0.,
	drop_path_rate=0., norm_layer=nn.LayerNorm, **kwargs):

but it is [mel_bins, time] when the model is defined in the ASIT_wrapper.py?

PretrainedSED/models/asit/ASIT_wrapper.py

Lines 10 to 16 in 1aa47e4

	self.asit = vit_base(
	patch_size=[16, 16],
	audio_size=[128, 592],
	stride=[16, 16],
	in_chans=1,
	num_classes=0
	)

I would expect that the order of these dimensions is fixed, since the positional encodings will change otherwise.