From 981fb7a22b95db79d6692c3a3071850b0e552d93 Mon Sep 17 00:00:00 2001 From: wimh966 Date: Sat, 21 Oct 2023 22:37:42 +0200 Subject: [PATCH] fix typos --- README.md | 6 +++--- exp/opt/int4_group.yaml | 4 ++-- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/README.md b/README.md index 26b8912..5a619d3 100644 --- a/README.md +++ b/README.md @@ -4,7 +4,7 @@ Official PyTorch implementation of O ## Overview -The Outlier Suppression+ (OS+) effectively suppresses outliers in large language models for better quantization performance without extra inference burden. It first identifies the outlier asymmetric shape across channels and proposes a channel-wise shifting technique with a migration pattern to eliminate it. It then focuses on the outlier concentration phenomenon and proposes to scale down outlier channels toward a dedicated objective. +The Outlier Suppression+ (OS+) effectively suppresses outliers in large language models for better quantization performance without extra inference burden. It first identifies the outlier asymmetric shape across channels and proposes a channel-wise shifting technique with a migration pattern to eliminate it. It then focuses on the outlier concentration phenomenon and proposes to scale down outlier channels toward an elaborate objective.

@@ -12,7 +12,7 @@ The Outlier Suppression+ (OS+) effectively suppresses outliers in large language We assess the efficacy of our approach under both standard and fine-grained quantization settings. On standard one, OS+ achieves near-floating-point performance on 8-bit and 6-bit BERT, OPTs, BLOOM, and BLOOMZ. On fine-grained one, OS+ can surpass others by 9.41\% on 4-bit LLaMA with per-token quantization and obtain lossless results on 4-bit OPT with per-group quantization. -In the following sections, [Support](#support) gives supported models and quantization schemes, [Getting Started](#getting started) introduces the whole procedure to run this project including data preparation, quantization, evaluation to updated model export. [Evaluation](#evaluation) lists configs for each table in the paper for others to reproduce. +In the following sections, [Support](#support) gives supported models and quantization schemes, [Getting Started](#getting started) introduces the whole procedure to run this project including data preparation, quantization, evaluation and updated model export. [Evaluation](#evaluation) lists configs for each table in the paper for other researchers to reproduce. ## Support @@ -146,7 +146,7 @@ quant: symmetric: True # True: symmetric quantization, False: asymmetric one ch_axis: -1 # 0: per-channel quantization, -1: per-layer one calibrate: 128 # calibration size - calibrate_path: /mnt/lustre/weixiuying.vendor/datasets/nlp_datasets/pile_cali # calibration path, make sure there is _cali in the name + calibrate_path: /mnt/lustre/weixiuying.vendor/datasets/nlp_datasets/pile_cali # calibration dataset path, make sure there is _cali in the name except_quantizer: null is_remove_padding: True # True: remove [PAD] during calibration migrate: True # True: shifting and scaling operations, False: no shifting and scaling operations. diff --git a/exp/opt/int4_group.yaml b/exp/opt/int4_group.yaml index 4fbe16d..73ad9c2 100644 --- a/exp/opt/int4_group.yaml +++ b/exp/opt/int4_group.yaml @@ -1,14 +1,14 @@ quant: a_qconfig: quantizer: GroupFixedFakeQuantize - group_size: 512 + group_size: 1024 observer: MinMaxObserver # EMAMSEObserver EMAMinMaxObserver EMAQuantileObserver EMAPruneMinMaxObserver bit: 4 symmetric: False ch_axis: 0 w_qconfig: quantizer: GroupFixedQuantize - group_size: 512 + group_size: 1024 observer: MinMaxObserver bit: 4 symmetric: False