EADST

Pytorch: Freeze layers to Finetune the Model

作者：XD / 发表： 2023年2月23日 22:02 / 更新： 2023年2月23日 22:03 / 编程笔记 / 阅读量：500

Pytorch: Freeze layers to Finetune the Model.

for k, v in model.named_parameters():
    print(k) # check the layer name
for k, v in model.named_parameters():
    if k in ["last.weight", "last.bias"]: # freeze the layer with the given name list
        v.requires_grad = True
    else:
        v.requires_grad = False

本文作者：XD 转载请标明出处：http://www.eadst.com/blog/166

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

本站现有博文242篇,共被浏览290904次

本站已经建立1780天!

热门文章
文章归档
回到顶部