AUTOROUND
AutoRound is an e-commerce platform for modified car parts.
AUTOROUND
Industry:
Automotive E-Commerce Internet
Founded:
2014-10-29
Address:
Shenzhen, Guangdong, China
Country:
China
Website Url:
http://www.autoround.com.cn
Total Employee:
11+
Status:
Active
Contact:
0755-86095699
Total Funding:
13 M CNY
Similar Organizations
Autozi.com
Autozi.com is an e-commerce platform that offers automobile parts.
Car10
Car10 is an e-commerce and digital automotive service platform.
Carzone
Carzone is an e-commerce platform for auto parts.
RumbleON
RumbleON provides an e-commerce platform that facilitates to sell, buy, trade pre-owned recreation vehicles.
Souche
Souche provides new and second-hand car trading services.
SpareXHub
SpareXHub is an e-commerce platform offering automotive spare parts.
Yottaa
Yottaa is an eCommerce acceleration platform that enables online brands to improve web performance and increase online conversion.
Current Employees Featured
Bokai Zhang Founder, Chairman & General Manager @ AutoRound
Founder, Chairman & General Manager
Jing Zhang CEO @ AutoRound
CEO
Qian Zhang COO @ AutoRound
COO
Founder
Investors List
QF Capital
QF Capital investment in Series A - AutoRound
Shidai jietong
Shidai jietong investment in Angel Round - AutoRound
Official Site Inspections
http://www.autoround.com.cn
- Host name: 120.24.234.238
- IP address: 120.24.234.238
- Location: Hangzhou China
- Latitude: 30.294
- Longitude: 120.1619
- Timezone: Asia/Shanghai
More informations about "AutoRound"
AutoRound - Crunchbase Company Profile & Funding
Organization. AutoRound . Connect to CRM . Save . Summary. Financials. People. Technology. Signals & News. Similar Companies. About. AutoRound is an e-commerce platform for …See details»
10 Tips for Quantizing LLMs and VLMs with AutoRound
Nov 22, 2024 2. Symmetric vs. Asymmetric Quantization. Suggestion: we recommend using full-range symmetric quantization, which AutoRound has set as the default. Background: While asymmetric quantization is ...See details»
QLoRA with AutoRound: Cheaper and Better LLM Fine-tuning on …
Oct 4, 2024 In this article, we will experiment and compare HQQ, AQLM, AutoRound, bitsandbytes, and GPTQ for QLoRA fine-tuning. We will see how fast they are for fine-tuning …See details»
The Impact of the Calibration Dataset for AutoRound and AWQ …
Oct 31, 2024 AutoRound appears even more resilient to the choice of calibration dataset, showing very similar performance across all English tasks. For French tasks, calibrating on …See details»
AutoRound: Accurate Low-bit Quantization for LLMs
Jun 29, 2024 AutoRound is also a weight-only quantization method. For AutoRound, Intel exploits SignSGD, an optimization method to find the optimal …See details»
auto-round - Oven
AutoRound is an advanced quantization algorithm for low-bits LLM/VLM inference. It's tailored for a wide range of models. AutoRound adopts sign gradient descent to fine-tune rounding values …See details»
LLM 量化算法AutoRound 0.3 发布及原理浅析 - 知乎 - 知乎专栏
AutoRound V0.3 特性. 支持了更多的设备:AutoRound格式支持CPU、HPU和CUDA推理，并且解决了2-bit kernel精度问题. 模型量化recipe:在 低比特开源LLM排行上发布了不少模型recipe， …See details»
AdaRound：一种自适应调整round的模型后量化算法
本文是对高通在2020年发表的论文《Up or Down? Adaptive Rounding for Post-Training Quantization》的学习和分享。 一、概要过往的模型量化的研究核心点都是如何寻一个更优scale值使得量化后的模型精度更高。而高…See details»
AutoRound：4比特量化的精度突破 - CSDN博客
Mar 8, 2025 文章浏览阅读971次，点赞13次，收藏30次。AutoRound是由Intel提出的一种基于SignSGD的低比特量化方法，旨在保持模型高精度的同时显著加快量化速度。该方法特别适用 …See details»
PeaceInMind-CSDN博客
Feb 28, 2024 在尽量公正的评估中，AutoRound在W4G128、W4G-1、W3G128、W2G128的大多数场景中优于GPTQ，AWQ等方法。 2024-02-28 10:05:10 1175 1 原创 语言模型1 …See details»
Weight Only Quantization (WOQ) 项目介绍 intel GPU 量化实现
Aug 29, 2024 RTN（Rounding to Nearest）：一种简单直观的方法，将值四舍五入为最近的整数。它不需要额外的数据集，量化速度快，适用于W4G32或W8配置，但在较低精度级别下效果 …See details»
Intel(R) Neural Compressor – Medium
Mar 1, 2024 The AutoRound Quantization Algorithm. Weight-Only Quantization for LLMs Across Hardware Platforms. Apr 2, 2024. 2. Apr 2, 2024. 2. Published in. Intel Analytics Software. Run …See details»
Low-Bit Quantized Open LLM Leaderboard - Intel
Table 1. Accuracy (higher is better) of fp16 7B vs. 13B AutoRound models. Moreover, AutoRound can accommodate the majority of popular models, whereas other quantization algorithms have …See details»
Microsoft-phi-4-Instruct-AutoRound-GPTQ-4bit - Hugging Face
Model Card for Microsoft-phi-4-Instruct-AutoRound-GPTQ-4bit Model Overview Model Name: Microsoft-phi-4-Instruct-AutoRound-GPTQ-4bit Model Type: Instruction-tuned, Quantized GPT …See details»
auto-round/README.md at main · intel/auto-round - GitHub
AutoRound is an advanced quantization algorithm for low-bits LLM/VLM inference. It's tailored for a wide range of models. AutoRound adopts sign gradient descent to fine-tune rounding values …See details»
The Best Quantization Methods to Run Llama 3.1 on Your GPU
Aug 12, 2024 It depends on what you want to do. It seems that there are no good reasons to choose AutoRound (asym) and GPTQ. We can forget them for Llama 3.1 8B. AutoRound …See details»
QLoRA with AutoRound: Cheaper and Better LLM Fine-tuning on
Aug 19, 2024 In this article, we will experiment and compare HQQ, AQLM, AutoRound, bitsandbytes, and GPTQ for QLoRA fine-tuning. We will see how fast they are for fine-tuning …See details»
China Wheel Hub Assembly, Tensioner Pulley, Timing Belt Kit ...
Hangzhou Duhui Automotive Parts Co.,Ltd: Find professional wheel hub assembly, tensioner pulley, timing belt kit, overrunning alternater pulley and hydraulic clutch release bearing …See details»
OPEA/QwQ-32B-int4-AutoRound-gptq-sym - Hugging Face
Mar 6, 2025 Model Details This model is an int4 model with group_size 128 and symmetric quantization of Qwen/QwQ-32B generated by intel/auto-round algorithm.. How To Use INT4 …See details»
Autoround | Walmart Canada
Storage & organization ... Autoround 513273 Front Wheel Hub And Bearing Assembly Compatible with 2008-2016 Dodge Grand Caravan 2008-2016 Chrysler Town & Country 2009-2014 VW …See details»