---
license: mit
base_model: roberta-large
tags:
- generated_from_trainer
datasets:
- fursov/gec_ner_val3
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: ner-gec-roberta-large-v4
  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: 0.643409688321442
    - name: Recall
      type: recall
      value: 0.5775246056357017
    - name: F1
      type: f1
      value: 0.6086894738711854
    - name: Accuracy
      type: accuracy
      value: 0.9614897122818877
---

<!-- 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-roberta-large-v4

This model is a fine-tuned version of [roberta-large](https://huggingface.co./roberta-large) on the fursov/gec_ner_val3 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2489
- Precision: 0.6434
- Recall: 0.5775
- F1: 0.6087
- Accuracy: 0.9615

## 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: 64
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10.0

### Training results

| Training Loss | Epoch | Step | Accuracy | F1     | Validation Loss | Precision | Recall |
|:-------------:|:-----:|:----:|:--------:|:------:|:---------------:|:---------:|:------:|
| 0.2536        | 0.58  | 500  | 0.9347   | 0.0376 | 0.2469          | 0.0814    | 0.0245 |
| 0.2316        | 1.15  | 1000 | 0.9359   | 0.1365 | 0.2339          | 0.2272    | 0.0975 |
| 0.2175        | 1.73  | 1500 | 0.9392   | 0.1823 | 0.2172          | 0.2842    | 0.1342 |
| 0.1757        | 2.3   | 2000 | 0.9438   | 0.3123 | 0.1979          | 0.4011    | 0.2556 |
| 0.1682        | 2.88  | 2500 | 0.9502   | 0.3911 | 0.1817          | 0.4787    | 0.3307 |
| 0.121         | 3.46  | 3000 | 0.9537   | 0.4504 | 0.1753          | 0.5310    | 0.3910 |
| 0.0982        | 4.03  | 3500 | 0.9556   | 0.4980 | 0.1807          | 0.5606    | 0.4480 |
| 0.0858        | 4.61  | 4000 | 0.9577   | 0.5304 | 0.1732          | 0.5867    | 0.4839 |
| 0.0563        | 5.18  | 4500 | 0.1839   | 0.6007 | 0.5155          | 0.5548    | 0.9585 |
| 0.0586        | 5.76  | 5000 | 0.1804   | 0.6231 | 0.5237          | 0.5691    | 0.9605 |
| 0.0404        | 6.34  | 5500 | 0.1948   | 0.6214 | 0.5423          | 0.5792    | 0.9599 |
| 0.0397        | 6.91  | 6000 | 0.1994   | 0.6309 | 0.5458          | 0.5852    | 0.9610 |
| 0.0281        | 7.49  | 6500 | 0.2131   | 0.6345 | 0.5568          | 0.5931    | 0.9610 |
| 0.0182        | 8.06  | 7000 | 0.2249   | 0.6507 | 0.5649          | 0.6047    | 0.9625 |
| 0.0188        | 8.64  | 7500 | 0.2322   | 0.6413 | 0.5782          | 0.6081    | 0.9612 |
| 0.0123        | 9.22  | 8000 | 0.2473   | 0.6506 | 0.5777          | 0.6120    | 0.9622 |
| 0.0123        | 9.79  | 8500 | 0.2491   | 0.6427 | 0.5771          | 0.6081    | 0.9614 |


### Framework versions

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