Yolov8 目标检测剪枝学习记录

最近在进行YOLOv8系列的轻量化，目前在网络结构方面的优化已经接近极限了，所以想要学习一下模型剪枝是否能够进一步优化模型的性能
这里主要参考了torch-pruning的基本使用，v8模型剪枝，Jetson nano部署剪枝YOLOv8
下面只是记录一个简单流程，用于后续使用在自己的任务和网络中，数据不作为参考

首先训练一个base模型用于参考

环境：Ultralytics YOLOv8.2.18 🚀 Python-3.10.14 torch-2.4.0 CUDA:0 (NVIDIA H100 PCIe, 81008MiB)

训练代码

参考网上或者自己写一个能训练即可，为了方便我将通用的记录下来，实测可用来自代码来源

from ultralytics import YOLO
import os

root = os.getcwd()
## 配置文件路径
name_yaml             = os.path.join(root, "ultralytics/datasets/VOC.yaml")
name_pretrain         = os.path.join(root, "yolov8s.pt")
## 原始训练路径
path_train            = os.path.join(root, "runs/detect/VOC")
name_train            = os.path.join(path_train, "weights/last.pt")
## 约束训练路径、剪枝模型文件
path_constraint_train = os.path.join(root, "runs/detect/VOC_Constraint")
name_prune_before     = os.path.join(path_constraint_train, "weights/last.pt")
name_prune_after      = os.path.join(path_constraint_train, "weights/last_prune.pt")
## 微调路径
path_fineturn         = os.path.join(root, "runs/detect/VOC_finetune")

def else_api():
    path_data = ""
    path_result = ""
    model = YOLO(name_pretrain) 
    metrics = model.val()  # evaluate model performance on the validation set
    model.export(format='onnx', opset=11, simplify=True, dynamic=False, imgsz=640)
    model.predict(path_data, device="0", save=True, show=False, save_txt=True, imgsz=[288,480], save_conf=True, name=path_result, iou=0.5)  # 这里的imgsz为高宽

def step1_train():
    model = YOLO(name_pretrain) 
    model.train(data=name_yaml, device="0,1", imgsz=640, epochs=50, batch=32, workers=16, save_period=1, name=path_train)  # train the model

## 2024.3.4添加【amp=False】
def step2_Constraint_train():
    model = YOLO(name_train) 
    model.train(data=name_yaml, device="0,1", imgsz=640, epochs=50, batch=32, amp=False, workers=16, save_period=1,name=path_constraint_train)  # train the model

def step3_pruning():
    from LL_pruning import do_pruning
    do_pruning(name_prune_before, name_prune_after)

def step4_finetune():
    model = YOLO(name_prune_after)     # load a pretrained model (recommended for training)
    model.train(data=name_yaml, device="0,1", imgsz=640, epochs=50, batch=32, workers=16, save_period=1, name=path_fineturn)  # train the model

step1_train()
# step2_Constraint_train()
# step3_pruning()
# step4_finetune()

第一步，step1_train()

即训练一个base模型，用于最后性能好坏的重要参考

第二步，step2_Constraint_train()

训练之前在ultralytics\engine\trainer.py添加bn的L1正则，使得bn参数在训练时变得稀疏

通过对参数的绝对值进行惩罚，使得一些不重要的权重变为零，从而实现模型的稀疏化和简化

     # Backward
     self.scaler.scale(self.loss).backward()
     ## add new code=============================duj
     ## add l1 regulation for step2_Constraint_train               
     l1_lambda = 1e-2 * (1 - 0.9 * epoch / self.epochs)
     for k, m in self.model.named_modules():
         if isinstance(m, nn.BatchNorm2d):
             m.weight.grad.data.add_(l1_lambda * torch.sign(m.weight.data))
             m.bias.grad.data.add_(1e-2 * torch.sign(m.bias.data))

     # Optimize - https://pytorch.org/docs/master/notes/amp_examples.html
     if ni - last_opt_step >= self.accumulate:
         self.optimizer_step()
         last_opt_step = ni

在这里插入图片描述

个人理解的稀疏化作用
- 通过对 gamma 和 beta 添加 L1 正则化，可以促使某些通道的 BN 权重变得非常小，甚至为零。这意味着在剪枝时，可以将这些通道从模型中移除
- 通过稀疏化 BN 层并剪除不重要的通道，剩下的通道会更有效地利用计算资源，减少无用计算。

第三步，step3_pruning()剪枝操作

LL_pruning.py

from ultralytics import YOLO
import torch
from ultralytics.nn.modules import Bottleneck, Conv, C2f, SPPF, Detect
import os


class PRUNE():
    def __init__(self) -> None:
        self.threshold = None

    def get_threshold(self, model, factor=0.8):
        ws = []
        bs = []
        for name, m in model.named_modules():
            if isinstance(m, torch.nn.BatchNorm2d):
                w = m.weight.abs().detach()
                b = m.bias.abs().detach()
                ws.append(w)
                bs.append(b)
                print(name, w.max().item(), w.min().item(), b.max().item(), b.min().item())
                print()
        # keep
        ws = torch.cat(ws)
        self.threshold = torch.sort(ws, descending=True)[0][int(len(ws) * factor)]

    def prune_conv(self, conv1: Conv, conv2: Conv):
        ## a. 根据BN中的参数，获取需要保留的index================
        gamma = conv1.bn.weight.data.detach()
        beta  = conv1.bn.bias.data.detach()
        
        keep_idxs = []
        local_threshold = self.threshold
        while len(keep_idxs) < 8:  ## 若剩余卷积核<8, 则降低阈值重新筛选
            keep_idxs = torch.where(gamma.abs() >= local_threshold)[0]
            local_threshold = local_threshold * 0.5
        n = len(keep_idxs)
        # n = max(int(len(idxs) * 0.8), p)
        print(n / len(gamma) * 100)
        # scale = len(idxs) / n

        ## b. 利用index对BN进行剪枝============================
        conv1.bn.weight.data = gamma[keep_idxs]
        conv1.bn.bias.data   = beta[keep_idxs]
        conv1.bn.running_var.data = conv1.bn.running_var.data[keep_idxs]
        conv1.bn.running_mean.data = conv1.bn.running_mean.data[keep_idxs]
        conv1.bn.num_features = n
        conv1.conv.weight.data = conv1.conv.weight.data[keep_idxs]
        conv1.conv.out_channels = n
        
        ## c. 利用index对conv1进行剪枝=========================
        if conv1.conv.bias is not None:
            conv1.conv.bias.data = conv1.conv.bias.data[keep_idxs]

        ## d. 利用index对conv2进行剪枝=========================
        if not isinstance(conv2, list):
            conv2 = [conv2]
        for item in conv2:
            if item is None: continue
            if isinstance(item, Conv):
                conv = item.conv
            else:
                conv = item
            conv.in_channels = n
            conv.weight.data = conv.weight.data[:, keep_idxs]
     
    def prune(self, m1, m2):
        if isinstance(m1, C2f):      # C2f as a top conv
            m1 = m1.cv2
        if not isinstance(m2, list): # m2 is just one module
            m2 = [m2]
        for i, item in enumerate(m2):
            if isinstance(item, C2f) or isinstance(item, SPPF):
                m2[i] = item.cv1
        self.prune_conv(m1, m2)
     
def do_pruning(modelpath, savepath):
    pruning = PRUNE()

    ### 0. 加载模型
    yolo = YOLO(modelpath)                  # build a new model from scratch
    pruning.get_threshold(yolo.model, 0.8)  # 获取剪枝时bn参数的阈值，这里的0.8为剪枝率。

    ### 1. 剪枝c2f 中的Bottleneck
    for name, m in yolo.model.named_modules():
        if isinstance(m, Bottleneck):
            pruning.prune_conv(m.cv1, m.cv2)

    ### 2. 指定剪枝不同模块之间的卷积核
    seq = yolo.model.model
    for i in [3,5,7,8]: 
        pruning.prune(seq[i], seq[i+1])

    ### 3. 对检测头进行剪枝
    # 在P3层: seq[15]之后的网络节点与其相连的有 seq[16]、detect.cv2[0] (box分支)、detect.cv3[0] (class分支)
    # 在P4层: seq[18]之后的网络节点与其相连的有 seq[19]、detect.cv2[1] 、detect.cv3[1] 
    # 在P5层: seq[21]之后的网络节点与其相连的有 detect.cv2[2] 、detect.cv3[2] 
    detect:Detect = seq[-1]
    last_inputs   = [seq[15], seq[18], seq[21]]
    colasts       = [seq[16], seq[19], None]
    for last_input, colast, cv2, cv3 in zip(last_inputs, colasts, detect.cv2, detect.cv3):
        pruning.prune(last_input, [colast, cv2[0], cv3[0]])
        pruning.prune(cv2[0], cv2[1])
        pruning.prune(cv2[1], cv2[2])
        pruning.prune(cv3[0], cv3[1])
        pruning.prune(cv3[1], cv3[2])

    ### 4. 模型梯度设置与保存
    for name, p in yolo.model.named_parameters():
        p.requires_grad = True
     
    yolo.val()
    torch.save(yolo.ckpt, savepath)
    yolo.model.pt_path = yolo.model.pt_path.replace("last.pt", os.path.basename(savepath))
    yolo.export(format="onnx")

    ## 重新load模型，修改保存命名，用以比较剪枝前后的onnx的大小
    yolo = YOLO(modelpath)  # build a new model from scratch
    yolo.export(format="onnx")


if __name__ == "__main__":

    modelpath = "runs/detect1/14_Constraint/weights/last.pt"
    savepath  = "runs/detect1/14_Constraint/weights/last_prune.pt"
    do_pruning(modelpath, savepath)

在这里插入图片描述

如下图可用看到剪枝前后还是有区别的，参数量减少很多，网络性能将不可用，需要微调恢复精度
查看剪枝前后模型大小 du -sh ./runs/detect/VOC_Constraint/weights/last*（yolov8n模型）

微调

该部分内容我也存在一些疑问，例如很多博主让ultralytics\engine\trainer.py添加加载模型代码，经过我8.2版本测试代码添加是完全失效的，因为setup_model在执行if isinstance(self.model, torch.nn.Module)就已经return了。

 def setup_model(self):
        """Load/create/download model for any task."""
        if isinstance(self.model, torch.nn.Module):  # if model is loaded beforehand. No setup needed
            return

例如ultralytics\engine\trainer.py
v8…x添加代码：548行参考这里

self.model = self.get_model(cfg=cfg, weights=weights, verbose=RANK == -1)
# duj add code to finetune
self.model = weights
return ckpt

如果是v8.0.x 参考这里

在看到这篇中的修改1启发

v8.2.x上面我不确定是哪个版本需要添加的，但是我实测都不起作用
我尝试在ultralytics\engine\model.py添加如下代码加载模型成功

 self.trainer = (trainer or self._smart_load("trainer"))(overrides=args, _callbacks=self.callbacks)
        if not args.get("resume"):  # manually set model only if not resuming
            # self.trainer.model = self.trainer.get_model(weights=self.model if self.ckpt else None, cfg=self.model.yaml)
            # self.model = self.trainer.model
            # dujiang edit 
            self.trainer.model = self.model.train()

            if SETTINGS["hub"] is True and not self.session: