EADST

Quick Review: AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration

作者：XD / 发表： 2023年12月7日 00:38 / 更新： 2023年12月7日 00:38 / 科研学习 / 阅读量：1149

AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration

Highlight:

  • Optimal Alpha Scaling: Focuses on determining the optimal alpha value for scaling weights prior to quantization.

本文作者：XD 转载请标明出处：http://www.eadst.com/blog/228

本站采用知识共享署名-非商业性使用-相同方式共享 4.0 国际许可协议进行许可。

上一篇
Quick Review: ZeroQuant-FP
下一篇
Quick Review: SmoothQuant: Accurate and Efficient Post-Training Quantization for LLMs
相关标签
LLM Quantization
About Me
XD
Goals determine what you are going to be.
Category
标签云
XGBoost C++ VGG-16 SQLite Zip PyCharm Google OpenCV hf YOLO Transformers Heatmap Hilton CEIR Excel SPIE Django Use XML SQL Vmess RGB NLP TensorFlow Disk Docker 腾讯云 Attention FP64 QWEN diffusers Knowledge Streamlit CTC Bipartite FP32 Paddle Sklearn Github VPN Permission CV Tiktoken HuggingFace Bitcoin MD5 Windows LeetCode llama.cpp LoRA PDF COCO Qwen Pillow Conda Shortcut TSV 算法题 FlashAttention Magnet EXCEL AI Tensor Ubuntu Land Video Hotel torchinfo scipy HaggingFace 域名 Pickle WebCrawler Linux Input Vim FastAPI Clash Food Translation Git GIT Qwen2.5 Jetson NLTK Safetensors Tracking Color Crawler GPT4 Website Random Pandas CAM ModelScope uwsgi Data Pytorch Base64 Hungarian Augmentation git Nginx Quantize 搞笑 Numpy 报税 Review Dataset OpenAI Interview PIP Cloudreve BeautifulSoup Anaconda CSV tar Password TensorRT GoogLeNet LaTeX OCR ResNet-50 Template Web SVR ChatGPT CUDA Bin GPTQ Diagram JSON Paper Gemma transformers 阿里云 Plate Bert Quantization 关于博主 v2ray Plotly Baidu FP16 Michelin 签证 logger Algorithm Datetime Python LLAMA Card git-lfs Image2Text DeepStream uWSGI ONNX Ptyhon Statistics BTC Markdown Domain Math BF16 LLM InvalidArgumentError Proxy GGML VSCode UNIX NameSilo Distillation
站点统计

本站现有博文266篇,共被浏览440771次

本站已经建立2019天!

热门文章
文章归档
回到顶部