metadata
language: en
tags:
- token-classification
- ner
- named-entity-recognition
- roberta
- restaurant
- mit-restaurant
datasets:
- mit_restaurant
metrics:
- f1
- precision
- recall
- accuracy
widget:
- text: I want a reservation at an italian restaurant with outdoor seating
example_title: Restaurant query
- text: Find me a cheap chinese place near downtown
example_title: Restaurant search
RoBERTa Large for MIT Restaurant NER
This model is a fine-tuned version of RoBERTa Large on the MIT Restaurant dataset for Named Entity Recognition (NER).
Model Description
- Model type: Token Classification (NER)
- Base model: roberta-large
- Dataset: MIT Restaurant NER dataset
- Languages: English
- Task: Named Entity Recognition for restaurant domain
Entity Types
The model can identify the following entity types: ['O', 'B-Amenity', 'I-Amenity', 'B-Cuisine', 'I-Cuisine', 'B-Dish', 'I-Dish', 'B-Hours', 'I-Hours', 'B-Location', 'I-Location', 'B-Price', 'I-Price', 'B-Rating', 'I-Rating', 'B-Restaurant_Name', 'I-Restaurant_Name']
Usage
from transformers import AutoTokenizer, AutoModelForTokenClassification, pipeline
tokenizer = AutoTokenizer.from_pretrained("niruthiha/roberta-large-mit-restaurant-ner")
model = AutoModelForTokenClassification.from_pretrained("niruthiha/roberta-large-mit-restaurant-ner")
# Using pipeline
nlp = pipeline("ner", model=model, tokenizer=tokenizer, aggregation_strategy="simple")
result = nlp("I want a reservation at an italian restaurant with outdoor seating")
print(result)
# Manual usage
inputs = tokenizer("I want a reservation at an italian restaurant", return_tensors="pt")
outputs = model(**inputs)
Training Details
- Fine-tuned on MIT Restaurant NER dataset
- Training epochs: 5
- Learning rate: 1e-5
- Batch size: 16
- Gradient accumulation steps: 2
Dataset
The MIT Restaurant dataset contains restaurant-related queries with entity annotations. Dataset source: https://groups.csail.mit.edu/sls/downloads/restaurant/
Performance
The model achieves good performance on restaurant domain NER tasks. Specific metrics will be updated after evaluation.