azif17632467/FannOrFlop

Fann Or Flop: A Multigenre, Multiera Benchmark for Arabic Poetry Understanding

Fann or Flop is the first comprehensive benchmark designed to evaluate large language models (LLMs) on their ability to understand Arabic poetry. It contains nearly 7,000 poem-explanation pairs covering 12 poetic eras, 21 genres, and multiple meters, providing a culturally rich and linguistically challenging testbed for Arabic NLP.

---

Latest Updates

🔥🔥 [20 Aug 2025] 🔥🔥 Fann or Flop accepted to EMNLP 2025 main track.
🔥 [26 May 2025] Fann or Flop the 1^st benchmark for assessing the LLM's ability to comprehend and analyze Arabic poetry is released.
🤗 [19 Feb 2025] Fann or Flop dataset available on Hugging Face.

---

Key Features

• Expert-Annotated Explanations: Verse-level commentary verified by native Arabic scholars.
• 12 Historical Eras: From Pre-Islamic and Umayyad to Modern poetry.
• Multi-Dimensional Evaluation: Faithfulness, fluency, metaphor, historical context, and rhetorical awareness.
• Structured Taxonomy: Each poem tagged with meter, genre, and era.
• QA-Style Format: Ideal for generative and comprehension-based evaluation in LLMs.

---

Dataset Summary

• Name: Fann or Flop
• Language: Arabic
• Samples: 6,984 poem–explanation pairs
• Task: Explanation generation, comprehension, QA-style evaluation
• Annotation Level: Verse-level and poem-level explanations
• Genres: مدح, هجاء, رثاء, غزل, etc.
• Eras Covered: Pre-Islamic to Modern (e.g., Jahiliyyah, Abbasid, Ottoman, Contemporary)
• Poetic Meters: الكامل, الطويل, البسيط, free verse, etc.

---

Dataset Structure

Each entry in the dataset contains:

Field	Type	Description
id	string	Unique poem identifier
title	string	Title of the poem
author	string	Name of the poet
source	string	URL to original poem
tags	list[str]	Meter, genre, and era (e.g., "الكامل", "مدح", "العصر الحديث")
meter	string	Poetic meter (e.g., الكامل, الطويل)
genre	string	Poetic genre (e.g., مدح, هجاء)
era	string	Historical era of the poem
verse_count	int	Number of verses
poem_verses	string	Full poem text (formatted with verse numbers)
explanation	list[dict]	List of dictionaries, each containing a verse and its detailed explanation
raw_explanation	string	Full poem explanation in paragraph format

---

Tasks and Use Cases

Fann or Flop can be used for a wide range of tasks, including:

• Poetic Explanation Generation (LLM text generation)
• Cultural and Historical QA (question answering from classical content)
• Verse-Level Comprehension
• Metrical & Stylistic Classification
• Cultural Understanding Evaluation

---

Evaluation & Metrics

• BLEU / chrF(++): Lexical overlap
• BERTScore: Semantic similarity (AraBERT, etc.)
• Textual Entailment: Consistency (mDeBERTa)
• Human Evaluation: 0–10 scale scoring:
• Literal understanding
• Thematic/emotional depth
• Cultural grounding
• Stylistic sensitivity
• Coherence and clarity

---

Model Benchmark Comparison on Fann or Flop

Model	BLEU	chrF(++)	BERTScore	Textual Entailment	Faithfulness / Consistency	Fluency / Grammaticality	Interpretive Depth
Closed Models
GPT-4o-2024-08-06 (OpenAI, 2024)	0.0395	0.2882	0.6410	0.6775	3.92 (± 0.99)	4.96 (± 0.20)	7.52
GPT-4o-mini-2024-07-18 (OpenAI, 2024)	0.0395	0.2542	0.6124	0.4383	2.91 (± 0.75)	4.28 (± 0.57)	7.50
Gemini-2.5-Flash (AI, 2025b)	0.0153	0.2618	0.6319	0.7475	4.25 (± 1.00)	4.98 (± 0.16)	7.22
Gemini-2.0-Flash (AI, 2025a)	0.0395	0.2618	0.6393	0.7154	3.99 (± 1.04)	4.95 (± 0.22)	6.50
Gemini-1.5-Pro (Reid et al., 2024)	0.0395	0.2618	0.6333	0.6180	3.59 (± 1.00)	4.80 (± 0.41)	5.38
Fanar-Star (Team et al., 2025)	0.0138	0.1538	0.5677	0.6468	2.16 (± 0.92)	3.40 (± 0.76)	2.88
Open Models
Deepseek-V3 (Liu et al., 2024)	0.0395	0.2771	0.6335	0.5117	3.36 (± 0.91)	4.98 (± 0.16)	4.75
Deepseek-R1 (Guo et al., 2025)	0.0395	0.2771	0.6335	0.5117	3.38 (± 0.92)	4.98 (± 0.16)	4.25
Llama-3.3-70B (Meta AI, 2024)	0.0153	0.2618	0.6393	0.5364	2.51 (± 0.90)	3.37 (± 0.73)	7.20
Qwen-3 (Team, 2025)	0.0296	0.2837	0.6158	0.6468	3.98 (± 0.90)	4.73 (± 0.45)	6.50
Aya-Expanse (Dang et al., 2024)	0.0329	0.2771	0.6328	0.6468	3.76 (± 0.90)	4.68 (± 0.47)	5.88
Jais (Sengupta et al., 2023)	0.0312	0.2698	0.6245	0.6023	3.21 (± 0.88)	4.35 (± 0.52)	5.35
ALLaM-7B (Bari et al., 2024)	0.0119	0.0463	0.5375	0.5997	1.32 (± 0.62)	2.11 (± 0.89)	3.12
AceGPT-v2-70B-Chat (Huang et al., 2023)	0.0402	0.0412	0.5759	0.6061	2.52 (± 0.91)	3.46 (± 0.95)	4.12

---

Citation

If you use Fann Or Flop dataset in your research, please consider citing:

@misc{alghallabi2025fannflopmultigenremultiera,
      title={Fann or Flop: A Multigenre, Multiera Benchmark for Arabic Poetry Understanding in LLMs}, 
      author={Wafa Alghallabi and Ritesh Thawkar and Sara Ghaboura and Ketan More and Omkar Thawakar and Hisham Cholakkal and Salman Khan and Rao Muhammad Anwer},
      year={2025},
      eprint={2505.18152},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2505.18152}, 
}