@浙大疏锦行
# 提取权重数据 import numpy as np weight_data = {} for name, param in model.named_parameters(): if 'weight' in name: weight_data[name] = param.detach().cpu().numpy() # 可视化权重分布 fig, axes = plt.subplots(1, len(weight_data), figsize=(15, 5)) fig.suptitle('Weight Distribution of Layers') for i, (name, weights) in enumerate(weight_data.items()): # 展平权重张量为一维数组 weights_flat = weights.flatten() # 绘制直方图 axes[i].hist(weights_flat, bins=50, alpha=0.7) axes[i].set_title(name) axes[i].set_xlabel('Weight Value') axes[i].set_ylabel('Frequency') axes[i].grid(True, linestyle='--', alpha=0.7) plt.tight_layout() plt.subplots_adjust(top=0.85) plt.show() # 计算并打印每层权重的统计信息 print("\n=== 权重统计信息 ===") for name, weights in weight_data.items(): mean = np.mean(weights) std = np.std(weights) min_val = np.min(weights) max_val = np.max(weights) print(f"{name}:") print(f" 均值: {mean:.6f}") print(f" 标准差: {std:.6f}") print(f" 最小值: {min_val:.6f}") print(f" 最大值: {max_val:.6f}") print("-" * 30)# 提取权重数据 import numpy as np weight_data = {} for name, param in model.named_parameters(): if 'weight' in name: weight_data[name] = param.detach().cpu().numpy() # 可视化权重分布 fig, axes = plt.subplots(1, len(weight_data), figsize=(15, 5)) fig.suptitle('Weight Distribution of Layers') for i, (name, weights) in enumerate(weight_data.items()): # 展平权重张量为一维数组 weights_flat = weights.flatten() # 绘制直方图 axes[i].hist(weights_flat, bins=50, alpha=0.7) axes[i].set_title(name) axes[i].set_xlabel('Weight Value') axes[i].set_ylabel('Frequency') axes[i].grid(True, linestyle='--', alpha=0.7) plt.tight_layout() plt.subplots_adjust(top=0.85) plt.show() # 计算并打印每层权重的统计信息 print("\n=== 权重统计信息 ===") for name, weights in weight_data.items(): mean = np.mean(weights) std = np.std(weights) min_val = np.min(weights) max_val = np.max(weights) print(f"{name}:") print(f" 均值: {mean:.6f}") print(f" 标准差: {std:.6f}") print(f" 最小值: {min_val:.6f}") print(f" 最大值: {max_val:.6f}") print("-" * 30)from tqdm import tqdm # 先导入tqdm库 import time # 用于模拟耗时操作 # 创建一个总步数为10的进度条 with tqdm(total=10) as pbar: # pbar是进度条对象的变量名 # pbar 是 progress bar（进度条）的缩写，约定俗成的命名习惯。 for i in range(10): # 循环10次（对应进度条的10步） time.sleep(0.5) # 模拟每次循环耗时0.5秒 pbar.update(1) # 每次循环后，进度条前进1步
