主页

索引

模块索引

搜索页面

6.3.4. PEFT

• PEFT
  • GET STARTED
    • 安装
    • Quicktour
      • PeftConfig
      • PeftModel
      • Save and load a model
      • Easy loading with Auto classes
  • TASK GUIDES
    • Image classification using LoRA
      • Install dependencies
      • a helper function
      • Load and prepare a model
      • run inference
    • Prefix tuning for conditional generation
      • Setup
      • Load dataset
      • Preprocess dataset
      • Train model
      • Inference
    • Prompt tuning for causal language modeling
      • Setup
      • Load dataset
      • Train
      • Inference
    • Semantic segmentation using LoRA
      • Prepare datasets for training and evaluation
      • Load a base model
      • Wrap the base model as a PeftModel for LoRA training
    • P-tuning for sequence classification
      • Setup
      • Load dataset and metric
      • Train
      • Inference
  • CONCEPTUAL GUIDES
    • LoRA
      • Merge LoRA weights into the base model
      • Utils for LoRA
      • Common LoRA parameters in PEFT
    • Prompting
      • Prompt tuning
      • Prefix tuning
      • P-tuning
      • 与Lora的区别-fromGPT
    • IA3
  • 参考
• PEFT 0.13.0
  • GET STARTED
    • Quicktour
      • train with the Trainer class
      • Inference
  • PEFT method guides
    • IA3
  • Developer guides
    • Model merging
      • Merge method
      • Merging (IA)³ Models
    • Quantization
    • LoRA
      • Initialization
      • Optimizers
      • Merge LoRA weights into the base model
      • Load adapters
      • Inference with different LoRA adapters in the same batch
        • 注意事项
    • Custom models
      • Multilayer perceptron
    • Adapter injection
      • Creating a new PEFT model
      • Saving the model
      • Loading the model
      • 查看相关参数-fromGPT
    • Mixed adapter types
    • torch.compile
    • PEFT checkpoint format
      • PEFT files
      • Convert to PEFT format
  • Accelerate integrations
    • DeepSpeed
    • Fully Sharded Data Parallel
  • Conceptual guides
    • Adapters
      • Low-Rank Adaptation (LoRA)
      • Mixture of LoRA Experts (X-LoRA)
      • Low-Rank Hadamard Product (LoHa)
      • Low-Rank Kronecker Product (LoKr)
      • Orthogonal Finetuning (OFT)
      • Orthogonal Butterfly (BOFT)
      • Adaptive Low-Rank Adaptation (AdaLoRA)
      • Llama-Adapter
    • Soft prompts
      • Multitask prompt tuning
    • IA3
      • 优点
      • Common IA3 parameters in PEFT
  • 参考

主页

索引

模块索引

搜索页面