Distilbert Base Uncased Finetuned Ner

🚀 distilbert-base-uncased-finetuned-ner

This model is a fine - tuned version of distilbert-base-uncased on the conll2003 dataset. It can effectively perform token classification tasks, achieving high precision, recall, F1 - score, and accuracy on the evaluation set.

🚀 Quick Start

This model is ready to be used for token classification tasks. You can load it using the Hugging Face Transformers library and start making predictions.

✨ Features

• Fine - tuned on conll2003: Trained specifically on the conll2003 dataset, which makes it well - suited for named entity recognition (NER) tasks.
• High performance: Achieves excellent results in terms of loss, precision, recall, F1 - score, and accuracy on the evaluation set.

📦 Installation

To use this model, you need to install the necessary libraries. You can install them using the following command:

pip install transformers datasets torch tokenizers

💻 Usage Examples

Here is a simple example of using this model for token classification:

from transformers import AutoTokenizer, AutoModelForTokenClassification
import torch

# Load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained("distilbert-base-uncased-finetuned-ner")
model = AutoModelForTokenClassification.from_pretrained("distilbert-base-uncased-finetuned-ner")

# Input text
text = "Some sample text for NER task."
inputs = tokenizer(text, return_tensors="pt")

# Make predictions
with torch.no_grad():
    logits = model(**inputs).logits
predicted_token_class_ids = logits.argmax(-1)
predicted_tokens_classes = [model.config.id2label[t.item()] for t in predicted_token_class_ids[0]]

print(predicted_tokens_classes)

📚 Documentation

Model Performance

This model achieves the following results on the evaluation set:

• Loss: 0.0611
• Precision: 0.9272
• Recall: 0.9382
• F1: 0.9327
• Accuracy: 0.9843

Training and Evaluation Data

The model is trained and evaluated on the conll2003 dataset.

Training Procedure

Training Hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e - 05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
• lr_scheduler_type: linear
• num_epochs: 3

Training Results

Framework Versions

• Transformers 4.9.1
• Pytorch 1.9.0+cu102
• Datasets 1.10.2
• Tokenizers 0.10.3

📄 License

This model is licensed under the Apache 2.0 license.