GitHub Homepage

A model fine-tuned for sentiment analysis based on vinai/phobert-base.

Labels:

  • NEG: Negative
  • POS: Positive
  • NEU: Neutral

Dataset: 30K e-commerce reviews

Usage

import torch
from transformers import RobertaForSequenceClassification, AutoTokenizer

model = RobertaForSequenceClassification.from_pretrained("wonrax/phobert-base-vietnamese-sentiment")

tokenizer = AutoTokenizer.from_pretrained("wonrax/phobert-base-vietnamese-sentiment", use_fast=False)

# Just like PhoBERT: INPUT TEXT MUST BE ALREADY WORD-SEGMENTED!
sentence = 'Đây là mô_hình rất hay , phù_hợp với điều_kiện và như cầu của nhiều người .'  

input_ids = torch.tensor([tokenizer.encode(sentence)])

with torch.no_grad():
    out = model(input_ids)
    print(out.logits.softmax(dim=-1).tolist())
    # Output:
    # [[0.002, 0.988, 0.01]]
    #     ^      ^      ^
    #    NEG    POS    NEU
Downloads last month
213
Safetensors
Model size
135M params
Tensor type
I64
·
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.