Why those tensors are different?

what is the difference between those tensor’s shapes?

Tesnor tensor([[-6.3570e-01, -9.6387e-01,  9.9367e-01, -1.2536e+00,  1.1033e+00,
         -1.4880e+00, -5.9717e-01,  6.9657e-01, -2.1082e-01, -1.3269e-02,
          7.7475e-01,  8.2461e-01, -4.4550e-01, -3.7213e-01, -2.5947e-02,
          1.1908e+00,  9.8890e-01,  4.9845e-01, -6.2397e-01, -7.4751e-01,
         -3.7960e-01, -1.0479e+00,  9.1659e-01,  2.0286e-01, -3.0425e+00,
          8.9924e-01,  4.9533e-01,  1.2649e-01, -7.8557e-01, -1.9877e+00,
          1.5009e+00,  6.3702e-01,  6.0425e-01, -2.5740e-01, -8.3749e-01, ....  -9.3429e-01,  3.5320e-01,
         -1.0368e-01, -6.1338e-01]], grad_fn=<SelectBackward>


tensor([[0.1703, 0.1442, 0.0332,  ..., 0.1482, 0.0598, 0.1676]],

this is for two different models with the same image but does this different form structure of the model or can I control the shape ?