---
license: apache-2.0
base_model: bert-base-uncased
tags:
- generated_from_trainer
datasets:
- fursov/gec_ner_val3
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: ner-gec-v2
  results:
  - task:
      name: Token Classification
      type: token-classification
    dataset:
      name: fursov/gec_ner_val3
      type: fursov/gec_ner_val3
    metrics:
    - name: Precision
      type: precision
      value: 0.36697832554186144
    - name: Recall
      type: recall
      value: 0.23284346770931644
    - name: F1
      type: f1
      value: 0.2849129753361379
    - name: Accuracy
      type: accuracy
      value: 0.941991634627572
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# ner-gec-v2

This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the fursov/gec_ner_val3 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2067
- Precision: 0.3670
- Recall: 0.2328
- F1: 0.2849
- Accuracy: 0.9420

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 128
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3.0

### Training results

| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
| 0.2324        | 1.15  | 500  | 0.2359          | 0.2070    | 0.0883 | 0.1238 | 0.9353   |
| 0.1901        | 2.3   | 1000 | 0.2137          | 0.3467    | 0.2212 | 0.2701 | 0.9399   |


### Framework versions

- Transformers 4.36.2
- Pytorch 2.1.0+cu118
- Datasets 2.16.1
- Tokenizers 0.15.0