--- license: gemma library_name: transformers pipeline_tag: text-generation extra_gated_button_content: Acknowledge license tags: - rag - qa language: - ar - en model-index: - name: SILMA-Kashif-2B-Instruct-v1.0 results: - task: type: text-generation dataset: name: SILMA RAGQA Benchmark Dataset V1.0 type: silma-ai/silma-rag-qa-benchmark-v1.0 metrics: - name: SILMA RAGQA Benchmark Score type: Average of Exact Match, BLEU, ROUGE, and BERTScore. value: 0.347 source: name: SILMA RAGQA Benchmark url: https://huggingface.co/datasets/silma-ai/silma-rag-qa-benchmark-v1.0 - task: type: text-generation dataset: name: OALL (All) type: OALL/Open-Arabic-LLM-Leaderboard metrics: - name: acc_norm type: loglikelihood_acc_norm value: 44.61 source: name: Open Arabic LLM Leaderboard url: https://huggingface.co/spaces/OALL/Open-Arabic-LLM-Leaderboard metrics: - bleu - bertscore - rouge - exact_match --- ## SILMA Kashif v1.0 (The Arabic RAG Model) * **SILMA Kashif 2B Instruct v1.0** is the premier release within the SILMA Kashif Family of models, specifically designed for **RAG** (Retrieval-Augmented Generation) tasks * Kashif excels in a specific task, answering questions based on contextual pieces in both Arabic and English. In addition, the model is also capable of performing Entity Extraction tasks as a minor skill * SILMA Kashif 2B v1.0 stands out as the top-performing open model in RAG within the 3-9 billion parameter range based on our evaluations using [SILMA RAGQA Benchmark](https://huggingface.co/datasets/silma-ai/silma-rag-qa-benchmark-v1.0) * SILMA Kashif is built on the powerful foundational models of Google Gemma, merging their strengths to deliver unmatched performance for users * Kashif is an open-weight model, free to use in accordance with our open license * Finally, the model comes with a context length of 12k **Important note:** Kashif is a specialized model which should ONLY be used in RAG setups. If you are looking for a general purpose model, please refer to [SILMA 9B Instruct v1.0](https://huggingface.co/silma-ai/SILMA-9B-Instruct-v1.0) ## Model Skills and Capabilities The model underwent intensive training to master a wide range of tasks and excel in performance. - The ability to answer questions in Arabic and English - The ability to deal with short and long contexts - The ability to provide short and long answers effectively - The ability to answer complex numerical questions - The ability to answer questions based on tabular data - Answering multi-hop questions: The ability to answer a single question using pieces of data from multiple paragraphs - Negative rejection: The ability to identify and exclude inaccurate answers, and provide a more accurate statement such as "The answer cannot be found in the given context" - Multi-domains: The ability to answer questions based on texts from different fields such as finance, medical, legal, etc. - The ability to deal with ambiguous contexts - The ability to extract entities from text - Ability to deal with diverse and complex prompts ## Model Evaluation ![silma-ragqa-benchmark-colored-2.png](https://cdn-uploads.huggingface.co/production/uploads/63d7acf73130cadcaf827e84/ufxxuArISZTVhK5q7YzaI.png) |Dataset | exact_match | rouge1 | bleu | bertscore| |---|---|---|---|---| |ragbench-finqa-en-test | 0.000 | 0.587 | 0.321 | 0.760| |ragbench-tatqa-ar-test | 0.000 | 0.484 | 0.130 | 0.774| |ragbench-tatqa-en-test | 0.059 | 0.646 | 0.423 | 0.808| |rag-instruct-benchmark-tester-en | 0.370 | 0.683 | 0.196 | 0.791| |ragbench-expertqa-en-test |0.000 | 0.465 | 0.151 | 0.677| |ragbench-msmarco-ar-test |0.000 | 0.144 | 0.096 | 0.781| |sciq-ar-test |0.170 | 0.000 | 0.000 | 0.753| |ragbench-covidqa-en-test |0.020 | 0.521 | 0.242 | 0.734| |ragbench-emanual-ar-test |0.000 | 0.237 | 0.159 | 0.806| |ragbench-finqa-ar-test |0.000 | 0.377 | 0.109 | 0.780| |xquad-r-validation-en |0.120 | 0.326 | 0.041 | 0.603| |ragbench-emanual-en-test |0.000 | 0.565 | 0.288 | 0.722| |xquad-r-ar-validation |0.070 | 0.130 | 0.042 | 0.698| |boolq-ar-test |0.450 | 0.000 | 0.000 | 0.700| |ragbench-hotpotqa-en-test |0.060 | 0.732 | 0.503 | 0.837| |ragbench-covidqa-ar-test |0.000 | 0.179 | 0.104 | 0.783| |ragbench-msmarco-en-test |0.020 | 0.491 | 0.207 | 0.729| |### Benchmark Average Scores |0.079 | 0.386 | 0.177 | 0.749| SILMA RAG QA Benchmark Score: 0.3478 ## SILMA AI [silma.ai](https://silma.ai) is a leading GenAI startup that excels in building and tailoring cutting-edge Large Language Models (LLMs) and AI technologies for the Arabic language ### Usage Below we share some code snippets on how to get quickly started with running the model. First, install the Transformers library with: ```sh pip install -U transformers ``` Then, copy the snippet from the section below #### Running with the `pipeline` API ```python import torch from transformers import pipeline pipe = pipeline( "text-generation", model="silma-ai/SILMA-Kashif-2B-Instruct-v1.0", model_kwargs={"torch_dtype": torch.bfloat16}, device="cuda", # replace with "mps" to run on a Mac device ) messages = [ {"role": "user", "content": """ أجب على السؤال بناءً على السياق أدناه السياق: تشمل الاتفاقيات رسوم حمل سنوية ثابت قدها 30 مليون جنيه إسترليني للقنوات نظراً لأن كلاً من مزوديها قادرين على تأمين دفعات إضافية إذا ما حققت هذه القنوات أهدافاً متعلقةً بالأداء. لا يوجد حالياً ما يشير إلى ما إذا كان الاتفاق الجديد يشمل محتوىً إضافياً كالفيديو عند الطلب والدقة العالية ، كذلك الذي سبق أن قدمته بي سكاي بي. وقد وافقت كل من بي سكاي بي و فيرجين ميديا على إنهاء الدعاوى القضائية بالمحكمة العليا ضد بعضهما بشأن معاليم الحمل التي تخص قنواتهما الأساسية. السؤال: ماسم الشركة التي وافقت على إنهاء دعواها القضائية ضد بي سكاي بي بالمحكمة العليا؟ الإجابة: """}, ] outputs = pipe(messages, max_new_tokens=600) assistant_response = outputs[0]["generated_text"][-1]["content"].strip() print(assistant_response) ``` - Response: ```text فيرجين ميديا "وقد وافقت كل من بي سكاي بي و فيرجين ميديا على إنهاء الدعاوى القضائية بالمحكمة العليا ضد بعضهما بشأن معاليم الحمل التي تخص قنواتهما الأساسية." ``` Note: for advanced usage examples such as multi-gpu, quantization or chat templates, please refer to [SILMA v1.0](https://huggingface.co/silma-ai/SILMA-9B-Instruct-v1.0#running-the-model-on-a-single--multi-gpu) examples ### Runing with Ollama ```sh ollama run hf.co/silma-ai/SILMA-Kashif-2B-Instruct-v1.0-GGUF ``` ### Prompt Format Here is a recommended way to prompt the model. You can modify the prompt based on your specific requirements, but if you encounter any challenges, following the format below in which we used to train the model may be helpful. - Arabic ```text أجب على السؤال بناءً على السياق أدناه السياق: ..... ..... السؤال: ... الإجابة: ... ``` - English ```text Answer the following question using the provided context below Context: ..... ..... Question: ... Answer: ... ``` ### GPU Requirements The following are the minimum/recommended GPU requirements for running inference: * Recommended * At least one GPU with a minimum of 24 GB of GPU memory * Examples: Nvidia RTX 4090 * Minimum * At least one GPU with 8 GB of GPU memory * Examples: Nvidia RTX 3070, RTX 3080 or T4 ## Effect of Quantization We have seen 2.6% drop in score (to 0.338) for the same model quantized 4bit ### Citation ```none @article{silma_01_2025, title={SILMA Kashif 2B Instruct v1.0}, url={https://huggingface.co/silma-ai/SILMA-Kashif-2B-Instruct-v1.0}, publisher={SILMA AI}, author={Silma Team}, year={2025} } ``` ### Intended Usage * The model should only be used in question answering use-cases such as RAG * The model can also be used to extract entities from text ### Limitations * Because it has few parameters, we've noticed that the model isn't very effective for handling complex numerical and financial reasoning, such as solving tricky calculations * The model has been trained specifically for text-based question answering, which may limit its ability to perform tasks beyond this scope, including simple tasks