OmniNER2024

Acknowledgments

Our code utilizes the framework from the LLaMA-Factory. We extend our gratitude to the contributors of this project. Additionally, we declare that the code we use adheres to the original project's copyright and licensing agreements.

Getting Started

Installation

git clone --depth 1 https://github.com/hiyouga/LLaMA-Factory.git
cd LLaMA-Factory
pip install -e ".[torch,metrics]"

Using OmniNER2024

To use the OmniNER2024 benchmark with the LLaMA-Factory framework, follow these steps:

1. Update data/dataset_info.json: Add dataset information for OmniNER2024 to the data/dataset_info.json file. Ensure the format and details are consistent with existing entries.

2. Add YAML Files: Add the YAML files from this repository (e.g., examples/train_lora/qwen2_7b_instruct_OmniNER_benchmarks/qwen7b_lora_sft.yaml) to the examples/train_lora directory in the llama-factory framework. Ensure the paths and dataset names in the YAML files match the entries in data/dataset_info.json.

Note: The OmniNER2024 benchmark is not yet open-sourced. However, we will be open-sourcing it soon and will update the repository with the benchmark once it is available.

Quickstart

Use the following commands to run LoRA fine-tuning and prediction for the Qwen2-7B-Instruct model, respectively.

Fine-tuning

llamafactory-cli train examples/train_lora/qwen7b_lora_sft.yaml

Prediction

llamafactory-cli train examples/train_lora/qwen7b_lora_predict.yaml

Evaluation

To evaluate the model on the OmniNER2024 benchmark, use the evaluate.py script. Here is how you can run the evaluation:

python examples/evaluate.py --path <path_to_predicted_jsonl> 

Replace <path_to_predicted_jsonl> with the path to the JSONL file generated from the prediction step.

OmniNER2024 using BERT

For instructions on using OmniNER2024 with BERT, please refer to the examples in the examples/BERT-NER directory in this repository. This directory contains detailed examples and scripts for fine-tuning and evaluating BERT models on the OmniNER2024 dataset.