Summary of Benchmarking the Performance Of Pre-trained Llms Across Urdu Nlp Tasks, by Munief Hassan Tahir et al.
Benchmarking the Performance of Pre-trained LLMs across Urdu NLP Tasks
by Munief Hassan Tahir, Sana Shams, Layba Fiaz, Farah Adeeba, Sarmad Hussain
First submitted to arxiv on: 24 May 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The study examines the performance of seven prominent Large Language Models (LLMs) across 17 tasks using 22 datasets and 13.8 hours of speech in a zero-shot setting. The LLMs include GPT-3.5-turbo, Llama 2-7B-Chat, Llama 3.1-8B, Bloomz 3B, Bloomz 7B1, Ministral-8B, and Whisper (Large, medium, and small variants). The study finds that state-of-the-art models currently outperform encoder-decoder models in most Urdu NLP tasks under zero-shot settings. However, the results also suggest that LLMs with fewer parameters but richer language-specific data can surpass state-of-the-art models in certain tasks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This study looks at how well seven large language models work on different tasks without being specifically trained for those tasks. The models are tested on 17 tasks using 22 datasets and a lot of speech. The results show that the best current models are better than these big language models at most Urdu-language tasks, but there’s one model that does surprisingly well. |
Keywords
» Artificial intelligence » Encoder decoder » Gpt » Llama » Nlp » Zero shot
