Spaces:
Runtime error
Runtime error
File size: 3,772 Bytes
617d510 d63d827 d1e02d2 617d510 c093cb0 6a6f9ef 617d510 7ccc2c1 617d510 6a6f9ef 617d510 27817dc 6a6f9ef 617d510 27817dc 6a6f9ef 617d510 27817dc 6a6f9ef b71697e 6a6f9ef 617d510 6a6f9ef 617d510 6a6f9ef 617d510 6a6f9ef |
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 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 |
import numpy as np
import pandas as pd
import torch
from torch import nn
from torch.utils.data import Dataset, DataLoader
from transformers import AutoTokenizer, BertModel
from sklearn import metrics
import streamlit as st
# Dataset for loading tables into DataLoader
class ToxicityDataset(Dataset):
def __init__(self, dataframe, tokenizer, max_len):
self.tokenizer = tokenizer
self.data = dataframe
self.text = self.data.comment_text
self.targets = self.data.labels
self.max_len = max_len
def __len__(self):
return len(self.text)
def __getitem__(self, index):
text = str(self.text[index])
text = " ".join(text.split())
inputs = self.tokenizer.encode_plus(
text,
None,
add_special_tokens=True,
max_length=self.max_len,
padding="max_length",
truncation=True,
return_token_type_ids=True,
)
ids = inputs["input_ids"]
mask = inputs["attention_mask"]
token_type_ids = inputs["token_type_ids"]
return {
"ids": torch.tensor(ids, dtype=torch.long),
"mask": torch.tensor(mask, dtype=torch.long),
"token_type_ids": torch.tensor(token_type_ids, dtype=torch.long),
"targets": torch.tensor(self.targets[index], dtype=torch.float),
}
# Freeze model and input tokens
def inference():
model.eval()
final_targets = []
final_outputs = []
with torch.no_grad():
for _, data in enumerate(testing_loader, 0):
ids = data["ids"].to(device, dtype=torch.long)
mask = data["mask"].to(device, dtype=torch.long)
token_type_ids = data["token_type_ids"].to(device, dtype=torch.long)
targets = data["targets"].to(device, dtype=torch.float)
outputs = model(ids, mask, token_type_ids)
final_targets.extend(targets.cpu().detach().numpy().tolist())
final_outputs.extend(torch.sigmoid(outputs).cpu().detach().numpy().tolist())
return final_outputs, final_targets
# Define models to be used
bert_path = "bert-base-uncased"
bert_tokenizer = AutoTokenizer.from_pretrained(bert_path)
bert_model = BertForSequenceClassification.from_pretrained(bert_path, num_labels=6)
tuned_model = model = torch.load("pytorch_bert_toxic.bin")
# Read and format data.
tweets_raw = pd.read_csv("test.csv", nrows=20)
labels_raw = pd.read_csv("test_labels.csv", nrows=20)
label_set = ["toxic", "severe_toxic", "obscene", "threat", "insult", "identity_hate"]
label_vector = labels_raw[label_set].values.tolist()
tweet_df = tweets_raw[["comment_text"]]
tweet_df["labels"] = label_vector
# User selects model for front-end.
option = st.selectbox("Select a text analysis model:", ("BERT", "Fine-tuned BERT"))
if option == "BERT":
tokenizer = bert_tokenizer
model = bert_model
else:
tokenizer = bert_tokenizer
model = tuned_model
# Based on user selection, prepare DataLoader
MAX_LENGTH = 100
TEST_BATCH_SIZE = 128
infer_dataset = ToxicityDataset(tweet_df, tokenizer, MAX_LENGTH)
infer_params = {"batch_size": TEST_BATCH_SIZE, "shuffle": True, "num_workers": 0}
testing_loader = DataLoader(test_dataset, **test_params)
prediction, targets = inference()
prediction = np.array(prediction) >= 0.5
targets = np.argmax(targets, axis=1)
prediction = np.argmax(prediction, axis=1)
accuracy = metrics.accuracy_score(targets, prediction)
f1_score_micro = metrics.f1_score(targets, prediction, average="micro")
f1_score_macro = metrics.f1_score(targets, prediction, average="macro")
st.write(prediction)
st.write(f"Accuracy Score = {accuracy}")
st.write(f"F1 Score (Micro) = {f1_score_micro}")
st.write(f"F1 Score (Macro) = {f1_score_macro}")
|