Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,117 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: mit
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
datasets:
|
| 4 |
+
- arbml/arabic_100k_reviews
|
| 5 |
+
language:
|
| 6 |
+
- ar
|
| 7 |
+
- en
|
| 8 |
+
base_model:
|
| 9 |
+
- google-bert/bert-base-uncased
|
| 10 |
+
pipeline_tag: text-classification
|
| 11 |
+
tags:
|
| 12 |
+
- fine-tuning-bert-arbic
|
| 13 |
+
- fine-tuning-bert-sentiment-analysis
|
| 14 |
+
- sentiment-analysis
|
| 15 |
+
- text-classification
|
| 16 |
+
- ktrain-library
|
| 17 |
+
---
|
| 18 |
+
Here's a modified README for your Hugging Face model page with the updated model name:
|
| 19 |
+
|
| 20 |
+
---
|
| 21 |
+
|
| 22 |
+
# Fine-Tuned Arabic Sentiment Analysis with BERT ๐
|
| 23 |
+
|
| 24 |
+
This repository contains a fine-tuned **BERT** model for sentiment analysis of Arabic reviews. The model is trained on the **[Arabic 100k Reviews](https://www.kaggle.com/datasets/abedkhooli/arabic-100k-reviews)** dataset and can classify reviews into three sentiment categories: **Positive**, **Negative**, and **Mixed**.
|
| 25 |
+
|
| 26 |
+
## Author ๐งโ๐ป
|
| 27 |
+
|
| 28 |
+
Khaled Soudy
|
| 29 |
+
GitHub: [khaledsoudy-1](https://github.com/khaledsoudy-1)
|
| 30 |
+
|
| 31 |
+
---
|
| 32 |
+
|
| 33 |
+
## How to Use the Model
|
| 34 |
+
|
| 35 |
+
### 1. Install Required Libraries
|
| 36 |
+
|
| 37 |
+
Make sure you have the **transformers** library installed:
|
| 38 |
+
|
| 39 |
+
```bash
|
| 40 |
+
pip install transformers
|
| 41 |
+
```
|
| 42 |
+
|
| 43 |
+
### 2. Load the Fine-Tuned Model
|
| 44 |
+
|
| 45 |
+
You can load the fine-tuned model and tokenizer directly from Hugging Face using the following code:
|
| 46 |
+
|
| 47 |
+
```python
|
| 48 |
+
from transformers import TFBertForSequenceClassification, BertTokenizer
|
| 49 |
+
|
| 50 |
+
# Load model and tokenizer from Hugging Face
|
| 51 |
+
model_name = "khaledsoudy/arabic-sentiment-bert-model"
|
| 52 |
+
|
| 53 |
+
# Load model
|
| 54 |
+
model = TFBertForSequenceClassification.from_pretrained(model_name)
|
| 55 |
+
|
| 56 |
+
# Load tokenizer
|
| 57 |
+
tokenizer = BertTokenizer.from_pretrained(model_name)
|
| 58 |
+
```
|
| 59 |
+
|
| 60 |
+
### 3. Use the Model for Prediction
|
| 61 |
+
|
| 62 |
+
To use the model for sentiment analysis on an Arabic text, follow these steps:
|
| 63 |
+
|
| 64 |
+
```python
|
| 65 |
+
# Sample Arabic text for sentiment prediction
|
| 66 |
+
text = "ุงูููุฏู ุฑุงุฆุน ู ุงูุฎุฏู
ุฉ ู
ู
ุชุงุฒุฉ"
|
| 67 |
+
|
| 68 |
+
# Tokenize the input text
|
| 69 |
+
inputs = tokenizer(text, return_tensors="tf")
|
| 70 |
+
|
| 71 |
+
# Get the model's prediction
|
| 72 |
+
outputs = model(**inputs)
|
| 73 |
+
|
| 74 |
+
# Get the predicted sentiment (assuming 3 classes: Positive, Negative, Mixed)
|
| 75 |
+
predicted_class = outputs.logits.argmax(axis=-1).numpy()[0]
|
| 76 |
+
|
| 77 |
+
# Map the predicted class index to sentiment labels
|
| 78 |
+
sentiment_labels = ['Positive', 'Negative', 'Mixed']
|
| 79 |
+
print(f"Predicted sentiment: {sentiment_labels[predicted_class]}")
|
| 80 |
+
```
|
| 81 |
+
|
| 82 |
+
### 4. Input Format
|
| 83 |
+
|
| 84 |
+
The model expects Arabic text input. The text should be preprocessed to remove unnecessary characters or diacritics for better results.
|
| 85 |
+
|
| 86 |
+
### 5. Sentiment Labels
|
| 87 |
+
|
| 88 |
+
The model classifies the sentiment into three categories:
|
| 89 |
+
|
| 90 |
+
- **Positive** ๐
|
| 91 |
+
- **Negative** ๐
|
| 92 |
+
- **Mixed** ๐ค
|
| 93 |
+
|
| 94 |
+
## Model Details
|
| 95 |
+
|
| 96 |
+
- **Model Name:** `khaledsoudy/arabic-sentiment-bert-model`
|
| 97 |
+
- **Model Type:** `TFBertForSequenceClassification`
|
| 98 |
+
- **Language:** Arabic
|
| 99 |
+
- **Sentiment Classes:** Positive, Negative, Mixed
|
| 100 |
+
|
| 101 |
+
## How to Fine-Tune This Model
|
| 102 |
+
|
| 103 |
+
You can fine-tune this model further using your own dataset. Check out the source code and related notebooks on my GitHub for detailed steps and guidance.
|
| 104 |
+
|
| 105 |
+
## License ๐
|
| 106 |
+
|
| 107 |
+
This model is licensed under the MIT License.
|
| 108 |
+
|
| 109 |
+
## Acknowledgments ๐
|
| 110 |
+
|
| 111 |
+
- **Hugging Face** for providing the platform to host models.
|
| 112 |
+
- **Google BERT** for the pre-trained model.
|
| 113 |
+
- **Kaggle** for the **Arabic 100k Reviews** dataset.
|
| 114 |
+
|
| 115 |
+
---
|
| 116 |
+
|
| 117 |
+
This README is ready for use on your Hugging Face model page! It includes detailed usage instructions, links to your GitHub, and other relevant information.
|