Model Card for Model ID

This model was developed by finetuning the DistilBERT Nepali Model. The model classifies the Nepali tweets related to COVID19 into three categories: neutral, positive and negative.

  • Developed by: Jeevan
  • Model type: DistilBERT Nepali
  • Language(s) (NLP): Nepali
  • Finetuned from model [optional]: DistilBERT Nepali Model

Training Details

Training Data

The dataset used for finetuning this model can be found at NepCOV19Tweets which contains Nepali tweets related to COVID-19.

Training HyperParameters

  • Batch size: 16
  • Learning Rate: 0.0001
  • Optimizer: AdamW
  • Epochs: 10

Evaluation

  • Training loss: 0.2414
  • Precision: 0.73
  • Recall: 0.73
  • F1 Score (Weighted): 0.73

Labels

  • Neutral: 0
  • Positive: 1
  • Negative: 2

USAGE

from transformers import pipeline

pipe = pipeline("text-classification", model="xap/Sentiment_Analysis_NepaliCovidTweets")
pipe("अमेरिकामा कोभिड बाट एकै दिन चार हजारभन्दा बढीको मृत्यु")

Citation

@misc {jeevan_2024,
    author       = { {jeevan} },
    title        = { Sentiment_Analysis_NepaliCovidTweets (Revision 3086409) },
    year         = 2024,
    url          = { https://huggingface.co/xap/Sentiment_Analysis_NepaliCovidTweets },
    doi          = { 10.57967/hf/2243 },
    publisher    = { Hugging Face }
}
Downloads last month
14
Safetensors
Model size
67M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.