SPEECH TRANSLATION USING LLMS using chatgpt.pptx

Durham University logo
Search
Open menu
Home
Study
Courses
Master of Laws M1K116
Master of Laws
Developed to reflect the vast range of legal contexts that exist in governmental structures across the globe, this course offers the flexibility to focus on your own areas of particular interest.
Alternative text
Start Dates
September 2024
Degree type
LLM

Course length
1 year full-time

Location
Durham City

Programme code
M1K116

Ready to Apply?
Navigate page
Course details
Learning
Entry requirements
Fees and funding
Career opportunities
Department information
Apply
Similar courses
Got any questions?
Share page





Person viewing the screen on a tablet
Get a Prospectus
Course details
Laws exist throughout the world, across continents and in different cultures and societies, governing the way we live, work and play. Our LLM Master of Laws reflects this variety of contexts and offers you the opportunity to choose from a range of legal specialisms, with the result that you can tailor your legal qualification to suit your interests.

You can choose from the widest range of modules, which can be selected from areas across the Law School including International Trade and Commercial Law, European Trade and Commercial Law, Corporate Law and International Law and Governance where subjects range from the law of the sea to electronic commerce.

This one-year full-time course is comprised of taught modules delivered through a mixture of lectures, small-group seminars or tutorials and film showings, as well as a major dissertation, which you will produce under supervision following independent research.

Students join the LLM from a broad range of countries and backgrounds, enriching the learning experience through the academic or professional experience they bring to the course. You will be able to participate in activities hosted by research centres who have their homes in the Law School, including the Institute for Commercial and Corporate Law, the Durham European Law Institute, the Centre for Criminal Law and Criminal Justice, Law and Global Justice at Durham and the Human Rights Centre.

Course Structure
Core modules
Applied Research Methods in Law provides specialised knowledge on the latest research methods and skills used in legal studies as well as an advanced understanding of their strengths and weaknesses and of their relevance for different forms of legal research. The module also creates a framework within which you will be able to critically assess potential research topics and, importantly, where you will be able to design, discuss and develop a detailed research proposal of the appropriate standards for your dissertation.

The Dissertation is based on an area of particular interest in an area of law and can be of one of two lengths – 10,000 or 15,000. Depending on the length chosen, the dissertation is equivalent to 2 or 2.5 modules.

To offer maximum flexibility, the remaining modules are chosen from an extensive range of options (* indica

1. SPEECH TRANSLATION USING LLM NAME- AMIT SINGH YADAV COURSE- PHD CSE ROLL NO- 2301201001 SUPERVISOR- DR. CHANDRESH KUMAR MAURYA

2. Table of Content Introduction Transformers BERT GPT Audio Conditioned LLM Speech Translation Conclusion References

3. INTRODUCTION Large Language Models (LLMs) are a class of artificial intelligence models designed to understand and generate human-like text. These models are typically based on deep learning architectures, particularly variants of recurrent neural networks (RNNs) or transformer architectures. Some examples of LLMS are GPT-3,ChatGPT,RoBERTa,XLNET etc.

4. TRANSFORMERS The core innovation of transformers is the self-attention mechanism. This mechanism allows the model to weigh the importance of different words in a sentence when encoding or decoding. It computes attention scores between all pairs of words in a sequence, capturing dependencies regardless of their position.

5. BERT BERT is based on the Transformer architecture, which is a neural network architecture specifically designed for sequence-to-sequence tasks, such as language translation and text generation .

6. GPT GPT is a generative model, meaning it can generate human-like text based on a given prompt or context. It's trained to predict the next word in a sequence given the preceding words, and this ability enables it to generate coherent and contextually relevant text.

7. HOW SPEECH TRANSLATION WORKS?

8. HOW SPEECH TRANSLATION WORKS?

9. CONCLUSION In conclusion, speech translation using LLMs offers a seamless and efficient solution for real-time multilingual communication, enabling individuals to overcome language barriers and interact effectively in diverse linguistic environments. This technology holds great potential for enhancing accessibility, fostering global collaboration, and facilitating cross-cultural communication in various domains.

10. Graves, A., Mohamed, A. R., & Hinton, G. (2013). Speech recognition with deep recurrent neural networks. In Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEE International Conference on (pp. 6645-6649). IEEE. Jimmy Lei Ba, Jamie Ryan Kiros , and Geoffrey E Hinton. Layer normalization. arXiv preprint arXiv:1607.06450, 2016. Vaswani, A., Shazeer , N., Parmar, N., Uszkoreit , J., Jones, L., Gomez, A. N., ... & Polosukhin , I. (2017). Attention is all you need. In Advances in neural information processing systems (pp. 5998-6008). Wang, Y., Skerry-Ryan, R., Stanton, D., Wu, Y., Jaitly , N., Yang, Z., ... & Shor, J. (2017). Tacotron : Towards end-to-end speech synthesis. arXiv preprint arXiv:1703.10135 . REFERENCES

11. THANK YOU