You can deploy this model and start earning money today!

nihad-ask/
Arabert-EOU-detection-model

Name: Arabert-EOU-detection-model
Availability: InStock

safetensors

bert

Views

No views yet

Arabic End-of-Turn (EOU) Detection Model — AraBERT Fine-Tuned

This model fine-tunes AraBERT for detecting end-of-turn (EOU) boundaries in Arabic dialogue.
It predicts whether a given user message represents a continuation or an end of turn.

Repository: nihad-ask/Arabert-EOU-detection-model
Task: Binary End-of-Utterance Classification
Language: Arabic (MSA + Dialects)
Base Model: aubmindlab/bert-base-arabertv2

🚦 Task Definition

This is a binary classification task:

Label	Meaning
0	Speaker will continue (NOT end of turn)
1	End of turn (EOU detected)

📌 Use Cases

Conversational AI / Chatbots
Dialogue Systems
Turn-taking prediction
Speech-to-text segmentation
Customer support automation

📊 Evaluation

Balanced Validation Set

Accuracy: 0.9539

Class	Precision	Recall	F1-score	Support
0 – Continue	0.9494	0.9589	0.9541	1702
1 – End of Turn	0.9585	0.9489	0.9536	1702

Overall:

Metric	Score
Accuracy	0.9539
Macro Avg F1	0.9539
Weighted Avg F1	0.9539
Total Samples	3404

Test Set

Accuracy: 0.8919

Class	Precision	Recall	F1-score	Support
0 – Continue	0.7671	0.9445	0.8466	3097
1 – End of Turn	0.9713	0.8676	0.9165	6705

Overall:

Metric	Score
Accuracy	0.8919
Macro Avg F1	0.8815
Weighted Avg F1	0.8944
Total Samples	9802

python

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

model_name = "nihad-ask/Arabert-EOU-detection-model"

tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)

text = "تمام و بعدين؟"

inputs = tokenizer(text, return_tensors="pt")
outputs = model(**inputs)
prediction = torch.argmax(outputs.logits, dim=1).item()

if prediction == 1:
    print("End of turn")
else:
    print("Speaker will continue")