---
library_name: transformers
license: apache-2.0
base_model: CAMeL-Lab/bert-base-arabic-camelbert-da-sentiment
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
model-index:
- name: egyptian_sentiment_analysis
  results: []
datasets:
- ehab215/egyptian_sentiment_analysis_dataset
language:
- en
- ar
pipeline_tag: text-classification
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# egyptian_sentiment_analysis

This model is a fine-tuned version of [CAMeL-Lab/bert-base-arabic-camelbert-da-sentiment](https://huggingface.co/CAMeL-Lab/bert-base-arabic-camelbert-da-sentiment) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2481
- Accuracy: 0.9519
- F1: 0.9520

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 5

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1     |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| No log        | 1.0   | 291  | 0.1533          | 0.9467   | 0.9466 |
| 0.2224        | 2.0   | 582  | 0.2004          | 0.9467   | 0.9469 |
| 0.2224        | 3.0   | 873  | 0.2178          | 0.9553   | 0.9553 |
| 0.0393        | 4.0   | 1164 | 0.2400          | 0.9553   | 0.9552 |
| 0.0393        | 5.0   | 1455 | 0.2481          | 0.9519   | 0.9520 |


### Framework versions

- Transformers 4.47.1
- Pytorch 2.5.1+cu121
- Tokenizers 0.21.0

### How to use:
```python
from transformers import AutoTokenizer, AutoModelForSequenceClassification
model_path = "ehab215/egyptian_sentiment_analysis"
tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForSequenceClassification.from_pretrained(model_path)

# Ensure model is on GPU if available
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device)

# Step 2: Prepare test examples
examples = [
    add any examples you would
]

# Tokenize the examples
inputs = tokenizer(examples, truncation=True, padding=True, return_tensors="pt", max_length=256)
inputs = {key: val.to(device) for key, val in inputs.items()}

# Step 3: Make predictions
with torch.no_grad():
    outputs = model(**inputs)
    logits = outputs.logits
    predictions = torch.argmax(logits, dim=-1).cpu().numpy()

# Step 4: Interpret results
label_map = {0: "negative", 1: "neutral", 2: "positive"}
predicted_labels = [label_map[p] for p in predictions]

# Display results
for text, label in zip(examples, predicted_labels):
    print(f"Text: {text}")
    print(f"Predicted Sentiment: {label}")
    print("-" * 50)
```