Upload Arabic Semantic Highlighter model

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

LICENSE

@@ -0,0 +1,80 @@
+ARABIC SEMANTIC HIGHLIGHTER MODEL LICENSE
+Non-Commercial Use Only
+
+Copyright (c) 2026 Hesham Haroon
+All rights reserved.
+
+1. DEFINITIONS
+
+"Model" refers to the Arabic Semantic Highlighter model, including all weights,
+configurations, tokenizer files, and associated documentation.
+
+"Commercial Use" means any use intended for or directed toward commercial
+advantage or monetary compensation, including but not limited to:
+- Using the Model in products or services sold for profit
+- Using the Model to provide paid services to third parties
+- Using the Model in applications that generate revenue
+- Incorporating the Model into proprietary software for sale
+
+"Non-Commercial Use" means use for research, education, personal projects,
+or other purposes not primarily intended for commercial advantage.
+
+2. GRANT OF LICENSE
+
+Subject to the terms of this License, the copyright holder grants you a
+worldwide, royalty-free, non-exclusive license to:
+
+a) Use the Model for Non-Commercial purposes
+b) Modify and create derivative works for Non-Commercial purposes
+c) Distribute the Model or derivative works for Non-Commercial purposes
+d) Use the Model for academic research and publication
+
+3. RESTRICTIONS
+
+You may NOT:
+
+a) Use the Model for any Commercial purpose without prior written permission
+b) Sell, lease, or rent the Model or any derivative works
+c) Use the Model in any product or service offered for monetary compensation
+d) Remove or alter any copyright notices or this License
+e) Use the Model in ways that violate applicable laws or regulations
+f) Use the Model to generate harmful, misleading, or illegal content
+
+4. ATTRIBUTION
+
+Any use of the Model must include appropriate attribution to the original
+author. When publishing research or other works that use this Model, please
+cite as specified in the README.md file.
+
+5. COMMERCIAL LICENSING
+
+For Commercial Use licensing inquiries, please contact the copyright holder
+to negotiate terms. Commercial licenses may be available on a case-by-case basis.
+
+6. NO WARRANTY
+
+THE MODEL IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE MODEL OR THE USE OR OTHER DEALINGS IN THE
+MODEL.
+
+7. TERMINATION
+
+This License automatically terminates if you violate any of its terms.
+Upon termination, you must cease all use of the Model and destroy any copies
+in your possession.
+
+8. GOVERNING LAW
+
+This License shall be governed by and construed in accordance with
+international copyright law.
+
+---
+
+For questions about this license or to request Commercial Use permissions,
+please contact: [Your Contact Information]
+
+Last updated: 2026

README.md

@@ -0,0 +1,162 @@
+# Arabic Semantic Highlighter
+
+A sentence-level semantic highlighting model for Arabic text, designed for RAG (Retrieval-Augmented Generation) systems.
+
+## Model Description
+
+This model identifies and highlights sentences in Arabic text that are relevant to a given query. It was fine-tuned on the [HeshamHaroon/arabic-semantic-relevance](https://huggingface.co/datasets/HeshamHaroon/arabic-semantic-relevance) dataset using span annotations.
+
+### Model Details
+
+- **Base Model:** BAAI/bge-reranker-base
+- **Task:** Sentence-level semantic relevance classification
+- **Language:** Arabic (العربية)
+- **Training Data:** ~66,000 query-sentence pairs extracted from span annotations
+
+### Performance Metrics
+
+| Metric | Score |
+|--------|-------|
+| Accuracy | 93.13% |
+| F1 Score | 94.58% |
+| Precision | 94.85% |
+| Recall | 94.30% |
+| AUC-ROC | 98.24% |
+
+## Usage
+
+```python
+import torch
+import numpy as np
+import re
+from transformers import AutoModelForSequenceClassification, AutoTokenizer
+
+class ArabicSemanticHighlighter:
+    def __init__(self, model_path):
+        self.model = AutoModelForSequenceClassification.from_pretrained(
+            model_path,
+            num_labels=1,
+        )
+        self.tokenizer = AutoTokenizer.from_pretrained(model_path)
+        self.device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+        self.model.to(self.device)
+        self.model.eval()
+
+    def _split_sentences(self, text, language="ar"):
+        if language == "ar":
+            sentences = re.split(r'[.؟!。\n]', text)
+        else:
+            sentences = re.split(r'[.?!\n]', text)
+        return [s.strip() for s in sentences if s.strip() and len(s.strip()) > 5]
+
+    def _score_sentence(self, question, sentence):
+        inputs = self.tokenizer(
+            question, sentence,
+            truncation=True,
+            max_length=256,
+            padding='max_length',
+            return_tensors='pt'
+        ).to(self.device)
+
+        with torch.no_grad():
+            logit = self.model(**inputs).logits.squeeze().item()
+            return 1 / (1 + np.exp(-logit))
+
+    def process(self, question, context, threshold=0.5, language="auto", return_sentence_metrics=False):
+        """
+        Highlight relevant sentences in context based on the question.
+
+        Args:
+            question: Query string
+            context: Text to search for relevant sentences
+            threshold: Minimum probability for relevance (default: 0.5)
+            language: "ar", "en", or "auto"
+            return_sentence_metrics: Include probability scores
+
+        Returns:
+            dict with highlighted_sentences, all_sentences, and optionally sentence_probabilities
+        """
+        if language == "auto":
+            arabic_chars = len(re.findall(r'[\u0600-\u06FF]', context))
+            language = "ar" if arabic_chars > len(context) * 0.3 else "en"
+
+        sentences = self._split_sentences(context, language)
+        probabilities = []
+        highlighted = []
+
+        for sentence in sentences:
+            prob = self._score_sentence(question, sentence)
+            probabilities.append(prob)
+            if prob >= threshold:
+                highlighted.append(sentence)
+
+        result = {
+            "highlighted_sentences": highlighted,
+            "all_sentences": sentences,
+        }
+
+        if return_sentence_metrics:
+            result["sentence_probabilities"] = probabilities
+
+        return result
+
+# Load model
+highlighter = ArabicSemanticHighlighter("path/to/model")
+
+# Example usage
+question = "ما هي فوائد الذكاء الاصطناعي في التعليم؟"
+context = """الذكاء الاصطناعي يحدث ثورة في قطاع التعليم.
+يساعد الذكاء الاصطناعي المعلمين في تخصيص المحتوى التعليمي لكل طالب.
+الطقس اليوم مشمس ودافئ."""
+
+result = highlighter.process(
+    question=question,
+    context=context,
+    threshold=0.5,
+    return_sentence_metrics=True
+)
+
+print("Highlighted sentences:", result["highlighted_sentences"])
+# Output: Relevant sentences about AI in education (excludes weather sentence)
+```
+
+## Training Details
+
+- **Epochs:** 3
+- **Batch Size:** 8
+- **Learning Rate:** 2e-5
+- **Max Sequence Length:** 256
+- **Gradient Accumulation Steps:** 4
+- **Optimizer:** AdamW with weight decay 0.01
+- **Training Time:** ~73 minutes on NVIDIA RTX 5060
+
+## Use Cases
+
+- **RAG Systems:** Highlight relevant passages for LLM context
+- **Search Results:** Show users which parts of documents match their query
+- **Document QA:** Identify answer-containing sentences
+- **Content Filtering:** Extract relevant information from long documents
+
+## Limitations
+
+- Optimized for Arabic text; may work on other languages but not tested
+- Best performance on sentences 10-200 characters in length
+- Requires GPU for efficient inference on large documents
+
+## Citation
+
+If you use this model, please cite:
+
+```bibtex
+@misc{arabic-semantic-highlighter,
+  author = {Hesham Haroon},
+  title = {Arabic Semantic Highlighter},
+  year = {2026},
+  publisher = {HuggingFace},
+  howpublished = {\url{https://huggingface.co/HeshamHaroon/arabic-semantic-highlighter}}
+}
+```
+
+## License
+
+This model is released under a **Non-Commercial License**. See [LICENSE](LICENSE) for details.

config.json

@@ -0,0 +1,33 @@
+{
+  "architectures": [
+    "XLMRobertaForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "dtype": "float32",
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "LABEL_0"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "LABEL_0": 0
+  },
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "xlm-roberta",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "output_past": true,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.57.3",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 250002
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cfd398375f844acca5a5d5470a7358fbfc0bb3c711a0f05d11b810cb31757e5
+size 1112201932

modeling_highlighter.py

@@ -0,0 +1,100 @@
+
+import torch
+import numpy as np
+import re
+from transformers import AutoModelForSequenceClassification, AutoTokenizer, PreTrainedModel, PretrainedConfig
+
+class SemanticHighlighterConfig(PretrainedConfig):
+    model_type = "semantic_highlighter"
+
+    def __init__(self, base_model_name="BAAI/bge-reranker-base", **kwargs):
+        super().__init__(**kwargs)
+        self.base_model_name = base_model_name
+
+class SemanticHighlighter(PreTrainedModel):
+    """Arabic Semantic Highlighter - highlights relevant sentences given a query."""
+
+    config_class = SemanticHighlighterConfig
+
+    def __init__(self, config):
+        super().__init__(config)
+        self.model = AutoModelForSequenceClassification.from_pretrained(
+            config._name_or_path,
+            num_labels=1,
+            ignore_mismatched_sizes=True
+        )
+        self.tokenizer = AutoTokenizer.from_pretrained(config._name_or_path)
+        self.device_type = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+
+    def forward(self, input_ids, attention_mask=None, labels=None):
+        return self.model(input_ids=input_ids, attention_mask=attention_mask, labels=labels)
+
+    def _split_sentences(self, text, language="ar"):
+        """Split text into sentences."""
+        if language == "ar":
+            sentences = re.split(r'[.؟!。\n]', text)
+        else:
+            sentences = re.split(r'[.?!\n]', text)
+        return [s.strip() for s in sentences if s.strip() and len(s.strip()) > 5]
+
+    def _score_sentence(self, question, sentence):
+        """Score a single sentence against the question."""
+        inputs = self.tokenizer(
+            question, sentence,
+            truncation=True,
+            max_length=256,
+            padding='max_length',
+            return_tensors='pt'
+        ).to(self.device_type)
+
+        with torch.no_grad():
+            logit = self.model(**inputs).logits.squeeze().item()
+            return 1 / (1 + np.exp(-logit))
+
+    def process(self, question, context, threshold=0.5, language="auto", return_sentence_metrics=False):
+        """
+        Process question and context to highlight relevant sentences.
+
+        Args:
+            question: The query/question string
+            context: The context text to search for relevant sentences
+            threshold: Minimum probability to consider a sentence relevant (default: 0.5)
+            language: Language of the text ("ar", "en", or "auto" for detection)
+            return_sentence_metrics: If True, include sentence probabilities in output
+
+        Returns:
+            dict with keys:
+                - highlighted_sentences: List of sentences above threshold
+                - all_sentences: All sentences in the context
+                - sentence_probabilities: (if return_sentence_metrics=True) probability scores
+        """
+        self.model.to(self.device_type)
+        self.model.eval()
+
+        # Auto-detect language
+        if language == "auto":
+            arabic_chars = len(re.findall(r'[\u0600-\u06FF]', context))
+            language = "ar" if arabic_chars > len(context) * 0.3 else "en"
+
+        # Split into sentences
+        sentences = self._split_sentences(context, language)
+
+        # Score each sentence
+        probabilities = []
+        highlighted = []
+
+        for sentence in sentences:
+            prob = self._score_sentence(question, sentence)
+            probabilities.append(prob)
+            if prob >= threshold:
+                highlighted.append(sentence)
+
+        result = {
+            "highlighted_sentences": highlighted,
+            "all_sentences": sentences,
+        }
+
+        if return_sentence_metrics:
+            result["sentence_probabilities"] = probabilities
+
+        return result

sentencepiece.bpe.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
+size 5069051

special_tokens_map.json

@@ -0,0 +1,51 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "cls_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "<mask>",
+    "lstrip": true,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "sep_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14917dd757b81bc44d4af6b028367351702656670c1954e055dabdfcf21593cf
+size 17082798

tokenizer_config.json

@@ -0,0 +1,56 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "250001": {
+      "content": "<mask>",
+      "lstrip": true,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "XLMRobertaTokenizer",
+  "unk_token": "<unk>"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be0ad0c4baf7bbfafb4f0e5edee081009ea9f8611c7d422822c0245787f41b62
+size 5777

training_info.json

@@ -0,0 +1,22 @@
+{
+  "model_name": "BAAI/bge-reranker-base",
+  "dataset": "HeshamHaroon/arabic-semantic-relevance",
+  "task": "sentence-level-highlighting",
+  "epochs": 3,
+  "batch_size": 8,
+  "learning_rate": 2e-05,
+  "train_sentences": 88110,
+  "test_results": {
+    "eval_loss": 0.2000904530286789,
+    "eval_accuracy": 0.9313270046455262,
+    "eval_f1": 0.9457562220804084,
+    "eval_precision": 0.94848,
+    "eval_recall": 0.9430480432707604,
+    "eval_auc": 0.9823959128165132,
+    "eval_runtime": 14.9552,
+    "eval_samples_per_second": 331.055,
+    "eval_steps_per_second": 20.729,
+    "epoch": 3.0
+  },
+  "completed_at": "2026-01-13T14:39:05.018353"
+}