You understood that the model is using layers 3, 5, 7, and 11, according to the cascade_layoutlmv3.yaml file.
However, in the original text, it states: "We adopt the standard practice to extract single-scale features from different Transformer layers, such as layers 4, 6, 8, and 12 of the LayoutLMv3 base model."
Describe Model I am using LayoutLM3:
You understood that the model is using layers 3, 5, 7, and 11, according to the cascade_layoutlmv3.yaml file.
However, in the original text, it states: "We adopt the standard practice to extract single-scale features from different Transformer layers, such as layers 4, 6, 8, and 12 of the LayoutLMv3 base model."
Is there a reason for changing the layers?