31ee80d556
* Make forward pass work * More improvements * Remove unused imports * Remove timm dependency * Improve loss calculation of token classifier * Fix most tests * Add docs * Add model integration test * Make all tests pass * Add LayoutLMv3FeatureExtractor * Improve integration test + make fixup * Add example script * Fix style * Add LayoutLMv3Processor * Fix style * Add option to add visual labels * Make more tokenizer tests pass * Fix more tests * Make more tests pass * Fix bug and improve docs * Fix import of processors * Improve docstrings * Fix toctree and improve docs * Fix auto tokenizer * Move tests to model folder * Move tests to model folder * change default behavior add_prefix_space * add prefix space for fast * add_prefix_spcae set to True for Fast * no space before `unique_no_split` token * add test to hightligh special treatment of added tokens * fix `test_batch_encode_dynamic_overflowing` by building a long enough example * fix `test_full_tokenizer` with add_prefix_token * Fix tokenizer integration test * Make the code more readable * Add tests for LayoutLMv3Processor * Fix style * Add model to README and update init * Apply suggestions from code review * Replace asserts by value errors * Add suggestion by @ducviet00 * Add model to doc tests * Simplify script * Improve README * a step ahead to fix * Update pair_input_test * Make all tokenizer tests pass - phew * Make style * Add LayoutLMv3 to CI job * Fix auto mapping * Fix CI job name * Make all processor tests pass * Make tests of LayoutLMv2 and LayoutXLM consistent * Add copied from statements to fast tokenizer * Add copied from statements to slow tokenizer * Remove add_visual_labels attribute * Fix tests * Add link to notebooks * Improve docs of LayoutLMv3Processor * Fix reference to section Co-authored-by: SaulLu <lucilesaul.com@gmail.com> Co-authored-by: Niels Rogge <nielsrogge@Nielss-MacBook-Pro.local>
2.9 KiB
Token classification with LayoutLMv3 (PyTorch version)
This directory contains a script, run_funsd_cord.py, that can be used to fine-tune (or evaluate) LayoutLMv3 on form understanding datasets, such as FUNSD and CORD.
The script run_funsd_cord.py leverages the 🤗 Datasets library and the Trainer API. You can easily customize it to your needs.
Fine-tuning on FUNSD
Fine-tuning LayoutLMv3 for token classification on FUNSD can be done as follows:
python run_funsd_cord.py \
--model_name_or_path microsoft/layoutlmv3-base \
--dataset_name funsd \
--output_dir layoutlmv3-test \
--do_train \
--do_eval \
--max_steps 1000 \
--evaluation_strategy steps \
--eval_steps 100 \
--learning_rate 1e-5 \
--load_best_model_at_end \
--metric_for_best_model "eval_f1" \
--push_to_hub \
--push_to_hub°model_id layoutlmv3-finetuned-funsd
👀 The resulting model can be found here: https://huggingface.co/nielsr/layoutlmv3-finetuned-funsd. By specifying the push_to_hub flag, the model gets uploaded automatically to the hub (regularly), together with a model card, which includes metrics such as precision, recall and F1. Note that you can easily update the model card, as it's just a README file of the respective repo on the hub.
There's also the "Training metrics" tab, which shows Tensorboard logs over the course of training. Pretty neat, huh?
Fine-tuning on CORD
Fine-tuning LayoutLMv3 for token classification on CORD can be done as follows:
python run_funsd_cord.py \
--model_name_or_path microsoft/layoutlmv3-base \
--dataset_name cord \
--output_dir layoutlmv3-test \
--do_train \
--do_eval \
--max_steps 1000 \
--evaluation_strategy steps \
--eval_steps 100 \
--learning_rate 5e-5 \
--load_best_model_at_end \
--metric_for_best_model "eval_f1" \
--push_to_hub \
--push_to_hub°model_id layoutlmv3-finetuned-cord
👀 The resulting model can be found here: https://huggingface.co/nielsr/layoutlmv3-finetuned-cord. Note that a model card gets generated automatically in case you specify the push_to_hub flag.