Chavo11

I did this as a workaround: embedding_vectors = test_outputs['layer1'] # randomly select d dimension condition = idx < embedding_vectors.shape[1] idx_1 = idx[condition] embedding_vectors = torch.index_select(embedding_vectors, 1, idx_1) for layer_name in...

@GlassyWing Can I ask you if you're refering to this particular implementation or to the paper itself?