File size: 10,883 Bytes
5abbc08 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 |
import torch
import torch.nn as nn
import numpy as np
import os
import json
from tqdm import tqdm
class time_travel_saver:
"""可视化数据提取器
用于保存模型训练过程中的各种数据，包括：
1. 模型权重 (.pth)
2. 高维特征 (representation/*.npy)
3. 预测结果 (prediction/*.npy)
4. 标签数据 (label/labels.npy)
"""
def __init__(self, model, dataloader, device, save_dir, model_name,
auto_save_embedding=False, layer_name=None,show = False):
"""初始化
Args:
model: 要保存的模型实例
dataloader: 数据加载器（必须是顺序加载的）
device: 计算设备(cpu or gpu)
save_dir: 保存根目录
model_name: 模型名称
"""
self.model = model
self.dataloader = dataloader
self.device = device
self.save_dir = save_dir
self.model_name = model_name
self.auto_save = auto_save_embedding
self.layer_name = layer_name
if show and not layer_name:
layer_dimensions = self.show_dimensions()
# print(layer_dimensions)
def show_dimensions(self):
"""显示模型中所有层的名称和对应的维度
这个函数会输出模型中所有层的名称和它们的输出维度，
帮助用户选择合适的层来提取特征。
Returns:
layer_dimensions: 包含层名称和维度的字典
"""
activation = {}
layer_dimensions = {}
def get_activation(name):
def hook(model, input, output):
activation[name] = output.detach()
return hook
# 注册钩子到所有层
handles = []
for name, module in self.model.named_modules():
if isinstance(module, nn.Module) and not isinstance(module, nn.ModuleList) and not isinstance(module, nn.ModuleDict):
handles.append(module.register_forward_hook(get_activation(name)))
self.model.eval()
with torch.no_grad():
# 获取一个batch来分析每层的输出维度
inputs, _ = next(iter(self.dataloader))
inputs = inputs.to(self.device)
_ = self.model(inputs)
# 分析所有层的输出维度
print("\n模型各层的名称和维度:")
print("-" * 50)
print(f"{'层名称':<40} {'特征维度':<15} {'输出形状'}")
print("-" * 50)
for name, feat in activation.items():
if feat is None:
continue
# 获取特征维度（展平后）
feat_dim = feat.view(feat.size(0), -1).size(1)
layer_dimensions[name] = feat_dim
# 打印层信息
shape_str = str(list(feat.shape))
print(f"{name:<40} {feat_dim:<15} {shape_str}")
print("-" * 50)
print("注: 特征维度是将输出张量展平后的维度大小")
print("你可以通过修改time_travel_saver的layer_name参数来选择不同的层")
print("例如：layer_name='avg_pool'或layer_name='layer4'等")
# 移除所有钩子
for handle in handles:
handle.remove()
return layer_dimensions
def _extract_features_and_predictions(self):
"""提取特征和预测结果
Returns:
features: 高维特征 [样本数, 特征维度]
predictions: 预测结果 [样本数, 类别数]
"""
features = []
predictions = []
indices = []
activation = {}
def get_activation(name):
def hook(model, input, output):
# 只在需要时保存激活值，避免内存浪费
if name not in activation or activation[name] is None:
activation[name] = output.detach()
return hook
# 根据层的名称或维度来选择层
# 注册钩子到所有层
handles = []
for name, module in self.model.named_modules():
if isinstance(module, nn.Module) and not isinstance(module, nn.ModuleList) and not isinstance(module, nn.ModuleDict):
handles.append(module.register_forward_hook(get_activation(name)))
self.model.eval()
with torch.no_grad():
# 首先获取一个batch来分析每层的输出维度
inputs, _ = next(iter(self.dataloader))
inputs = inputs.to(self.device)
_ = self.model(inputs)
# 如果指定了层名，则直接使用该层
if self.layer_name is not None:
if self.layer_name not in activation:
raise ValueError(f"指定的层 {self.layer_name} 不存在于模型中")
feat = activation[self.layer_name]
if feat is None:
raise ValueError(f"指定的层 {self.layer_name} 没有输出特征")
suitable_layer_name = self.layer_name
suitable_dim = feat.view(feat.size(0), -1).size(1)
print(f"使用指定的特征层: {suitable_layer_name}, 特征维度: {suitable_dim}")
else:
# 找到维度在指定范围内的层
target_dim_range = (256, 2048)
suitable_layer_name = None
suitable_dim = None
# 分析所有层的输出维度
for name, feat in activation.items():
if feat is None:
continue
feat_dim = feat.view(feat.size(0), -1).size(1)
if target_dim_range[0] <= feat_dim <= target_dim_range[1]:
suitable_layer_name = name
suitable_dim = feat_dim
break
if suitable_layer_name is None:
raise ValueError("没有找到合适维度的特征层")
print(f"自动选择的特征层: {suitable_layer_name}, 特征维度: {suitable_dim}")
# 保存层信息
layer_info = {
'layer_id': suitable_layer_name,
'dim': suitable_dim
}
layer_info_path = os.path.join(os.path.dirname(self.save_dir), 'layer_info.json')
with open(layer_info_path, 'w') as f:
json.dump(layer_info, f)
# 清除第一次运行的激活值
activation.clear()
# 现在处理所有数据
for batch_idx, (inputs, _) in enumerate(tqdm(self.dataloader, desc="提取特征和预测结果")):
inputs = inputs.to(self.device)
outputs = self.model(inputs) # 获取预测结果
# 获取并处理特征
feat = activation[suitable_layer_name]
flat_features = torch.flatten(feat, start_dim=1)
features.append(flat_features.cpu().numpy())
predictions.append(outputs.cpu().numpy())
# 清除本次的激活值
activation.clear()
# 移除所有钩子
for handle in handles:
handle.remove()
if len(features) > 0:
features = np.vstack(features)
predictions = np.vstack(predictions)
return features, predictions
else:
return np.array([]), np.array([])
def save_lables_index(self, path):
"""保存标签数据和索引信息
Args:
path: 保存路径
"""
os.makedirs(path, exist_ok=True)
labels_path = os.path.join(path, 'labels.npy')
index_path = os.path.join(path, 'index.json')
# 尝试从不同的属性获取标签
try:
if hasattr(self.dataloader.dataset, 'targets'):
# CIFAR10/CIFAR100使用targets属性
labels = np.array(self.dataloader.dataset.targets)
elif hasattr(self.dataloader.dataset, 'labels'):
# 某些数据集使用labels属性
labels = np.array(self.dataloader.dataset.labels)
else:
# 如果上面的方法都不起作用，则从数据加载器中收集标签
labels = []
for _, batch_labels in self.dataloader:
labels.append(batch_labels.numpy())
labels = np.concatenate(labels)
# 保存标签数据
np.save(labels_path, labels)
print(f"标签数据已保存到 {labels_path}")
# 创建数据集索引
num_samples = len(labels)
indices = list(range(num_samples))
# 创建索引字典
index_dict = {
"train": list(range(50000)), # 所有数据默认为训练集
"test": list(range(50000, 60000)), # 测试集索引从50000到59999
"validation": [] # 初始为空
}
# 保存索引到JSON文件
with open(index_path, 'w') as f:
json.dump(index_dict, f, indent=4)
print(f"数据集索引已保存到 {index_path}")
except Exception as e:
print(f"保存标签和索引时出错: {e}")
def save_checkpoint_embeddings_predictions(self, model = None):
"""保存所有数据"""
if model is not None:
self.model = model
# 保存模型权重
os.makedirs(self.save_dir, exist_ok=True)
model_path = os.path.join(self.save_dir,'model.pth')
torch.save(self.model.state_dict(), model_path)
if self.auto_save:
# 提取并保存特征和预测结果
features, predictions = self._extract_features_and_predictions()
# 保存特征
np.save(os.path.join(self.save_dir, 'embeddings.npy'), features)
# 保存预测结果
np.save(os.path.join(self.save_dir, 'predictions.npy'), predictions)
print("\n保存了以下数据:")
print(f"- 模型权重: {model_path}")
print(f"- 特征向量: [样本数: {features.shape[0]}, 特征维度: {features.shape[1]}]")
print(f"- 预测结果: [样本数: {predictions.shape[0]}, 类别数: {predictions.shape[1]}]") |