Muhammad Bin Usman's picture

9 3 29

Muhammad Bin Usman

Muhammad2003

·

AI & ML interests

- Model Alignment (SFT / DPO / ORPO ) - Model Merging / Pruning / MoE + latest tecniques - Instruction tuning and Preference datasets curation - Evaluation

Organizations

upvoted 3 collections over 1 year ago

haiku

🌸 This is a collection of synthetic datasets built to help improve the ability of open language models to better write haikus through the use of DPO • 3 items • Updated Jun 21, 2024 • 6

Llama 3.1

This collection hosts the transformers and original repos of the Llama 3.1, Llama Guard 3 and Prompt Guard models • 11 items • Updated Dec 6, 2024 • 705

🔀 Query Router

Dynamic routing system for querying LLMs, boosting efficiency. See https://github.com/MuhammadBinUsman03/Query-Router for more details. • 3 items • Updated Jul 20, 2024 • 2

Muhammad2003 (Muhammad Bin Usman)

Muhammad Bin Usman's picture

9 3 29

Muhammad Bin Usman

Muhammad2003

·

AI & ML interests

- Model Alignment (SFT / DPO / ORPO ) - Model Merging / Pruning / MoE + latest tecniques - Instruction tuning and Preference datasets curation - Evaluation

Organizations

upvoted 3 collections over 1 year ago

haiku

🌸 This is a collection of synthetic datasets built to help improve the ability of open language models to better write haikus through the use of DPO • 3 items • Updated Jun 21, 2024 • 6

Llama 3.1

This collection hosts the transformers and original repos of the Llama 3.1, Llama Guard 3 and Prompt Guard models • 11 items • Updated Dec 6, 2024 • 705

🔀 Query Router

Dynamic routing system for querying LLMs, boosting efficiency. See https://github.com/MuhammadBinUsman03/Query-Router for more details. • 3 items • Updated Jul 20, 2024 • 2