| license: mit | |
| datasets: | |
| - conll2003 | |
| language: | |
| - en | |
| metrics: | |
| - f1 | |
| library_name: peft | |
| pipeline_tag: token-classification | |
| tags: | |
| - unsloth | |
| - llama-2 | |
| [<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="150"/>](https://github.com/unslothai/unsloth) | |
| At the moment of writing the 🤗 transformers library doesn't have a Llama implementation for Token Classification ([although there is a open PR](https://github.com/huggingface/transformers/pull/29878)). | |
| This model is based on a [implementation](https://github.com/huggingface/transformers/issues/26521#issuecomment-1868284434) by community member [@KoichiYasuoka](https://github.com/KoichiYasuoka). | |
| * Base Model: `unsloth/llama-2-7b-bnb-4bit` | |
| * LORA Model Adaptation with rank 8 and alpha 32, other adapter configurations can be found in [`adapter_config.json`](https://huggingface.co/SauravMaheshkar/unsloth-llama-2-7b-bnb-4bit-conll2003-rank-4/blob/main/adapter_config.json) | |
| This model was only trained for a single epoch, however a notebook is made available for those who want to train on other datasets for longer. |