| --- |
| language: fa |
| pipeline_tag: text-classification |
| library_name: transformers |
| --- |
| |
| # QomSSLab/SubjectClassifier-v1 |
|
|
| This repository hosts a sequence-classification head trained for text classification. |
|
|
| ## Usage |
|
|
| ```python |
| from transformers import AutoTokenizer, AutoModelForSequenceClassification, pipeline |
| |
| model_id = "QomSSLab/SubjectClassifier-v1" |
| tokenizer = AutoTokenizer.from_pretrained(model_id) |
| model = AutoModelForSequenceClassification.from_pretrained(model_id) |
| classifier = pipeline("text-classification", model=model, tokenizer=tokenizer) |
| |
| text = "مثال از یک ورودی فارسی" |
| print(classifier(text)) |
| ``` |
|
|
| ## Labels |
|
|
| - `اخلاق` |
| - `تاریخ` |
| - `حدیث` |
| - `سیاسی` |
| - `فقه و احکام` |
| - `قرآن` |
| - `مشاوره` |
| - `کلام و اعتقادات` |
|
|
| ## Metrics |
|
|
| ## Validation Metrics |
|
|
| - Precision: 0.9717 |
| - Recall: 0.9713 |
| - F1: 0.9713 |
| - Accuracy: 0.9713 |
|
|
| ### Per-label Breakdown |
|
|
| | Label | Precision | Recall | F1 | Support | |
| | --- | --- | --- | --- | --- | |
| | اخلاق | 0.9455 | 0.9598 | 0.9526 | 199 | |
| | تاریخ | 0.9815 | 1.0000 | 0.9907 | 53 | |
| | حدیث | 0.9487 | 0.9823 | 0.9652 | 113 | |
| | سیاسی | 1.0000 | 0.9693 | 0.9844 | 163 | |
| | فقه و احکام | 0.9652 | 1.0000 | 0.9823 | 222 | |
| | قرآن | 1.0000 | 0.9875 | 0.9937 | 80 | |
| | مشاوره | 0.9670 | 0.9263 | 0.9462 | 95 | |
| | کلام و اعتقادات | 0.9818 | 0.9558 | 0.9686 | 226 | |
|
|
|
|
|
|
