app.py

import numpy as np
import os
import gradio as gr

os.environ["WANDB_DISABLED"] = "true"

from datasets import load_dataset, load_metric
from transformers import (
    AutoConfig,
    AutoModelForSequenceClassification,
    AutoTokenizer,
    TrainingArguments,
    logging,
pipeline
)




# model_name = 


# tokenizer = AutoTokenizer.from_pretrained(model_name)

# config = AutoConfig.from_pretrained(model_name)

# pipe = pipeline("text-classification")

# pipe("This restaurant is awesome")




# Question answering pipeline, specifying the checkpoint identifier
model = AutoModelForSequenceClassification.from_pretrained(
    pretrained_model_name_or_path= "thak123/Cro-Frida",
    num_labels=3,
)


analyzer = pipeline(

    "sentiment-analysis", model=model, tokenizer="EMBEDDIA/crosloengual-bert"

)
def predict_sentiment(x):
    return analyzer(x)





interface = gr.Interface(
    fn=predict_sentiment,
    inputs='text',
    outputs=['label'],
    title='Latvian Twitter Sentiment Analysis',
    examples= ["Es mīlu Tevi","Es ienīstu kafiju"],
    description='Get the positive/neutral/negative sentiment for the given input.'
)

interface.launch(inline = False)