microsoft/layoutlmv3-base
Updated
•
1.7M
•
326
The LayoutLM series are Transformer encoders useful for document AI tasks such as invoice parsing, document image classification and DocVQA.
Note Currently the best LayoutLM model.
Note A multilingual variant trained on 100 languages.
Note A LayoutLM (v1) model fine-tuned to perform question answering over documents (DocVQA).
Note A LayoutLMv3 model fine-tuned on the FUNSD dataset, a benchmark for document parsing.