• Home
• People
• Jobs
• Visit
• Thesis Topics
• Publications
• Conference Rankings
• Research
• Bcast
• Teaching
• Contact

SPONSORS

COLLABORATIONS

The documents distributed by this server have been provided by the contributing authors as a means to ensure timely dissemination of scholarly and technical work on a noncommercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.

Publications of SPCL

S. Ashkboos, M. Nikdan, S. Tabesh, R. L. Castro, T. Hoefler, D. Alistarh:
	HALO: Hadamard-Assisted Lower-Precision Optimization for LLMs (In Proceedings of the Neural Information Processing Systems, presented in San Diego, CA, USA, Dec. 2025) Abstract Quantized training of Large Language Models (LLMs) remains an open challenge, as maintaining accuracy while performing all matrix multiplications in low precision has proven difficult. This is particularly the case when fine-tuning pre-trained models, which can have large weight and activation outlier values that make lower-precision optimization difficult. To address this, we present HALO, a novel quantization-aware training approach for Transformers that enables accurate and efficient low-precision training by combining 1) strategic placement of Hadamard rotations in both forward and backward passes, which mitigate outliers, 2) high-performance kernel support, and 3) FSDP integration for low-precision communication. Our approach ensures that all large matrix multiplications during the forward and backward passes are executed in lower precision. Applied to LLAMA-family models, HALO achieves near-full-precision-equivalent results during fine-tuning on various tasks, while delivering up to 1.41x end-to-end speedup for full fine-tuning on RTX 4090 GPUs. HALO efficiently supports both standard and parameterefficient fine-tuning (PEFT). Our results demonstrate the first practical approach to fully quantized LLM fine-tuning that maintains accuracy in 8-bit precision, while delivering performance benefits. Documents download article: access preprint on arxiv:
	BibTeX
@inproceedings{ashkboos2025halo,   author={Saleh Ashkboos and Mahdi Nikdan and Soroush Tabesh and Roberto L. Castro and Torsten Hoefler and Dan Alistarh},   title={{HALO: Hadamard-Assisted Lower-Precision Optimization for LLMs}},   year={2025},   month={12},   booktitle={Proceedings of the Neural Information Processing Systems},   location={San Diego, CA, USA}, }