计算模型中的可训练参数数量：print_trainable_parameters【参考LoRA】

def print_trainable_parameters(model):
    """
    Prints the number of trainable parameters in the model.
    """
    trainable_params = 0
    all_param = 0
    for _, param in model.named_parameters():
        all_param += param.numel()
        if param.requires_grad:
            trainable_params += param.numel()
    print(f"trainable params: {trainable_params} || all params: {all_param} || trainable%: {100 * trainable_params / all_param}")

运行一下：

print_trainable_parameters(model)

输出结果如下：

trainable params: 8388608 || all params: 6666862592 || trainable%: 0.12582542214183376

---

---

---

2023年的深度学习入门指南(12) - PEFT与LoRA_Jtag特工的博客-CSDN博客