metadata
language: en
datasets:
- botdevringring/EN_Sentiment_3label_Dataset
widget:
- text: The delivery man was late and my package was badly damaged
- text: obviously, to sign the contract
- text: Good service for a fair price. Friendly and attentive staff.
- text: Call directly to make an appointment during the week
pipeline_tag: text-classification
tags:
- DistilBert
- Sentiment
- Pytorch
metrics:
- Accuracy, F1 Score
model-index:
- name: botdevringring/en-naxai-ai-sentiment-classification-154012122023
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: sentiment
type: sentiment
config: default
split: test
metrics:
- type: accuracy
value: 0.886
name: Accuracy
verified: true
- type: loss
value: 0.337
name: loss
verified: true
- type: precision
value: 0.827
name: Precision Macro
verified: true
- type: precision
value: 0.886
name: Precision Micro
verified: true
- type: precision
value: 0.88
name: Precision Weighted
verified: true
- type: recall
value: 0.777
name: Recall Macro
verified: true
- type: recall
value: 0.886
name: Recall Micro
verified: true
- type: recall
value: 0.886
name: Recall Weighted
verified: true
- type: f1
value: 0.794
name: F1 Macro
verified: true
- type: f1
value: 0.886
name: F1 Micro
verified: true
- type: f1
value: 0.881
name: F1 Weighted
verified: true
- type: samples per second
value: 179.335
name: samples p/second
verified: true
- type: steps per second
value: 22.432
name: steps p/second
verified: true
- type: epochs
value: 7
name: epochs
verified: true
en-naxai-ai-sentiment-classification-154012122023(latest)
The model is trained on the sentiment classification task in English. It uses 3 labels: -1, 0 and 1. These labels represent negative, neutral and positive sentiments.
This model is finetuned from DistilBERT base model (uncased). This model is a distilled version of the BERT base model. This model is uncased: it does not make a difference between english and English.
Model Details
- Language: en
- Problem type: Multi-class Classification
- Model Architecture: Distilbert base uncased
- Model Name: en-naxai-ai-sentiment-classification-154012122023
- Creation date: 15:40h 12/12/2023
- CO2 Emissions (in grams): 0.06
Classification Report:
Label | Precision | Recall | f1-Score | Support |
---|---|---|---|---|
-1 | 0.94 | 0.94 | 0.94 | 4870 |
0 | 0.58 | 0.59 | 0.59 | 867 |
1 | 0.91 | 0.90 | 0.91 | 2856 |
How to use this model
You can use Python to access this model:
from transformers import pipeline
analyzer = pipeline(
task='text-classification',
model=“botdevringring/en-naxai-ai-sentiment-classification-154012122023”,
tokenizer="botdevringring/en-naxai-ai-sentiment-classification-154012122023"
)
result = analyzer(
"Good service for a fair price. Friendly and attentive staff."
)
result
[{'label': '2', 'score': 0.515792965888977}]
Python API:
from transformers import AutoModelForSequenceClassification, AutoTokenizer
model = AutoModelForSequenceClassification.from_pretrained("botdevringring/en-naxai-ai-sentiment-classification-154012122023", use_auth_token=True)
tokenizer = AutoTokenizer.from_pretrained("botdevringring/en-naxai-ai-sentiment-classification-154012122023", use_auth_token=True)
inputs = tokenizer("Good service for a fair price. Friendly and attentive staff.", return_tensors="pt")
outputs = model(**inputs)
Or you can use cURL:
curl https://api-inference.huggingface.co/models/botdevringring/en-naxai-ai-sentiment-classification-154012122023 \
-X POST \
-d '{"inputs": "Good service for a fair price. Friendly and attentive staff."}' \
-H 'Content-Type: application/json' \
-H "Authorization: Bearer <Your HF API token>"
Acknowledgements
Model trained by Eduardo Brigham for Naxai powered by The Ring Ring Company