Electra Small Ner

🚀 Electra-small NER Model

This is an electra-small model fine-tuned for NER prediction tasks, capable of predicting location, person, and organization entities.

🚀 Quick Start

Use the following code to start using the model:

from transformers import AutoModelForTokenClassification, AutoTokenizer
from transformers import pipeline

model = AutoModelForTokenClassification.from_pretrained("rv2307/electra-small-ner")
tokenizer = AutoTokenizer.from_pretrained("rv2307/electra-small-ner")

nlp = pipeline("ner",
              model=model,
              tokenizer=tokenizer,device="cpu",
              aggregation_strategy = "max")

✨ Features

• Entity Prediction: The model can predict three types of entities: Location, Person, and Organization.
• Fine-tuned Model: Fine-tuned from Google's electra small discriminator.

📚 Documentation

Model Details

Model Description

This model is an electra-small model fine-tuned for the NER prediction task. Currently, it predicts the following three entities:

1. Location
2. Person
3. Organization

• Developed by: விபின் (Vipin)
• Model type: Google's electra small discriminator
• Language(s) (NLP): English
• License: Apache 2.0
• Finetuned from model [optional]: Google's electra small discriminator

Model Sources [optional]

• Repository: https://huggingface.co/google/electra-small-discriminator

Uses

This model uses a tokenizer from the distilbert family. So, the model may predict wrong entities for the same word (different sub-words). Use 'aggregation_strategy' to "max" when using the transformer's pipeline. For example, for 'ashwin': - "ash" => Person - "win" => Location

Out-of-Scope Use

The model may not work well for some long sentences.

Training Details

Training Procedure

This model is trained for 6 epochs with a learning rate of 3e-4.

 [39168/39168 41:18, Epoch 6/6]
Step	Training Loss	Validation Loss	Precision	Recall	F1	Accuracy
10000	0.086300	0.088625	0.863476	0.876271	0.869827	0.972581
20000	0.059800	0.079611	0.894612	0.884521	0.889538	0.976563
30000	0.050400	0.074552	0.895812	0.902591	0.899188	0.978380

Evaluation

The validation loss of this model is 0.07.

📄 License

This model is licensed under the Apache 2.0 license.