OPTQ: Accurate post-training quantization for generative pre-trained transformers

conference paper OPTQ: Accurate post-training quantization for generative pre-trained transformers published yes Elias Frantar author 09a8f98d-ec99-11ea-ae11-c063a7b7fe5f Saleh Ashkboos author Torsten Hoefler author Dan-Adrian Alistarh author 4A899BFC-F248-11E8-B48F-1D18A9856A870000-0003-3650-940X DaAl department ICLR: International Conference on Learning Representations Elastic Coordination for Scalable Machine Learning project Generative Pre-trained Transformer models, known as GPT or OPT, set themselves apart through breakthrough performance across complex language modelling tasks, but also by their extremely high computational and storage costs. Specifically, due to their massive size, even inference for large, highly-accurate GPT models may require multiple performant GPUs, which limits the usability of such models. While there is emerging work on relieving this pressure via model compression, the applicability and performance of existing compression techniques is limited by the scale and complexity of GPT models. In this paper, we address this challenge, and propose OPTQ, a new one-shot weight quantization method based on approximate second-order information, that is both highly-accurate and highly-efficient. Specifically, OPTQ can quantize GPT models with 175 billion parameters in approximately four GPU hours, reducing the bitwidth down to 3 or 4 bits per weight, with negligible accuracy degradation relative to the uncompressed baseline. Our method more than doubles the compression gains relative to previously-proposed one-shot quantization methods, preserving accuracy, allowing us for the first time to execute an 175 billion-parameter model inside a single GPU for generative inference. Moreover, we also show that our method can still provide reasonable accuracy in the extreme quantization regime, in which weights are quantized to 2-bit or even ternary quantization levels. We show experimentally that these improvements can be leveraged for end-to-end inference speedups over FP16, of around 3.25x when using high-end GPUs (NVIDIA A100) and 4.5x when using more cost-effective ones (NVIDIA A6000). The implementation is available at https://github.com/IST-DASLab/gptq. https://research-explorer.ista.ac.at/download/17378/17385/2023_ICLR_Frantar.pdf application/pdfno International Conference on Learning Representations2023Kigali, Rwanda eng 11th International Conference on Learning Representations https://research-explorer.ista.ac.at/record/17485 https://github.com/IST-DASLab/gptq Frantar, Elias, et al. “OPTQ: Accurate Post-Training Quantization for Generative Pre-Trained Transformers.” 11th International Conference on Learning Representations , International Conference on Learning Representations, 2023. Frantar, E., Ashkboos, S., Hoefler, T., & Alistarh, D.-A. (2023). OPTQ: Accurate post-training quantization for generative pre-trained transformers. In 11th International Conference on Learning Representations . Kigali, Rwanda: International Conference on Learning Representations. E. Frantar, S. Ashkboos, T. Hoefler, D.-A. Alistarh, in:, 11th International Conference on Learning Representations , International Conference on Learning Representations, 2023. E. Frantar, S. Ashkboos, T. Hoefler, and D.-A. Alistarh, “OPTQ: Accurate post-training quantization for generative pre-trained transformers,” in 11th International Conference on Learning Representations , Kigali, Rwanda, 2023. Frantar, Elias, Saleh Ashkboos, Torsten Hoefler, and Dan-Adrian Alistarh. “OPTQ: Accurate Post-Training Quantization for Generative Pre-Trained Transformers.” In 11th International Conference on Learning Representations . International Conference on Learning Representations, 2023. Frantar E, Ashkboos S, Hoefler T, Alistarh D-A. OPTQ: Accurate post-training quantization for generative pre-trained transformers. In: 11th International Conference on Learning Representations . International Conference on Learning Representations; 2023. Frantar E, Ashkboos S, Hoefler T, Alistarh D-A. 2023. OPTQ: Accurate post-training quantization for generative pre-trained transformers. 11th International Conference on Learning Representations . ICLR: International Conference on Learning Representations. 173782024-08-04T22:01:22Z2026-07-29T13:48:39Z