Spaces:
Running
Running
File size: 2,371 Bytes
ab4f033 72b3984 ab4f033 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 |
# imports
from transformers import pipeline
import gradio as gr
import pandas as pd
# define nlp mask
model = "siebert/sentiment-roberta-large-english"
nlp = pipeline(model=model) # set device=0 to use GPU (CPU default, -1)
# perform inference on given file
def inference(df, filename):
# texts & ids
texts = df[df.columns[1]].to_list()
ids = df[df.columns[0]].to_list()
# create new df based on csv inputs
new_df = pd.DataFrame(columns=[df.columns[0], df.columns[1], "Label", "Score"])
# iterate over texts, perform inference
for index in range(len(texts)):
preds = nlp(texts[index])
pred_sentiment = preds[0]["label"]
pred_score = preds[0]["score"]
print(texts[index])
print(preds)
# write data into df
# predicted sentiment
new_df.at[index, "Label"] = pred_sentiment
# predicted score
new_df.at[index, "Score"] = pred_score
# write text
new_df.at[index, df.columns[1]] = texts[index]
# write ID
new_df.at[index, df.columns[0]] = ids[index]
# export new file
n_filename = filename.name.split(".")[0] + "_csiebert_sentiment.csv"
new_df.to_csv(n_filename, index=False)
# return new file
return n_filename
# handle file reading for both csv and excel files
def read_file(filename):
# check type of input file
if filename.name.split(".")[1] == "csv":
print("entered")
# read file, drop index if exists
df = pd.read_csv(filename.name, index_col=False)
# perform inference on given .csv file
result = inference(df=df, filename=filename)
print("computed")
return result
elif filename.name.split(".")[1] == "xlsx":
df = pd.read_excel(filename.name, index_col=False)
# handle Unnamed
if df.columns[0] == "Unnamed: 0":
df = df.drop("Unnamed: 0", axis=1)
# perform inference on given .xlsx file
result = inference(df=df, filename=filename)
return result
# if neither csv nor xlsx provided -> exit
else:
return
gr.Interface(read_file,
inputs=[gr.inputs.File(label="Input file")],
outputs=[gr.outputs.File(label="Output file")],
description="Sentiment analysis: Input a csv/xlsx of form ID, Text. App performs sentiment analysis on Texts and exports results as new csv to download.",
allow_flagging=False,
layout="horizontal",
).launch() |