from_pretrained

1. AutoTokenizer.from_pretrained：从path/vocab.json中加载tokenizer

2. AutoConfig.from_pretrained：从path/config.json中加载模型配置信息

   • 更新模型配置信息：model = Model(config)

3. PreTrainedModel.from_pretrained：加载模型结构和模型参数

load_checkpoint

从 checkpoint 中加载模型 parameter，而不加载模型结构

本栏目推荐文章
• 2024-01-13 Can't perform a React state update on an unmounted component. This is a no-op, but it indicates a memory leak in your application. ==》引用了未使用的方法导致
• 状态模式（state）
• xPU pipeline and state machine
• 状态模式 State
• prometheus监控k8s集群信息（通过kube-state-metrics）
• 关于再次出现 Could not open client transport with JDBC Uri: jdbc:hive2://node1:10000: java.net.ConnectException: 拒绝连接 (Connection refused) (state=08S01,code=0)
• 【论文阅读笔记】【多模态-Vision-Language Pretraining】 BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation
• 《Mamba: Linear-Time Sequence Modeling with Selective State Spaces》阅读笔记
• Linux: CPU C-states
• AM trader Change in State of Delivery (CISD) - Making Reversals Easy

PreTrainedModel from_pretrained load_state_dict pretrained statepretrainedmodel from_pretrained load_state_dict from_pretrained huggingface from_pretrained pretrained函数 load_state_dict load_state_dict strict state load 模型model load load_state_dict pretrainedmodel pretrain pretraining pretrained