如何将此张量流代码转换为pytorch代码？

import tensorflow as tf
import torch

tf.reset_default_graph()

# 定义输入张量
x = tf.placeholder(tf.float32, shape=(None, 784), name='x')
y = tf.placeholder(tf.int32, shape=(None,), name='y')

# 定义模型
hidden = tf.layers.dense(x, 256, activation=tf.nn.relu)
logits = tf.layers.dense(hidden, 10)

# 定义损失函数和优化器
loss = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(labels=y, logits=logits))
optimizer = tf.train.AdamOptimizer(learning_rate=0.001)
train_op = optimizer.minimize(loss)

sess = tf.Session()
sess.run(tf.global_variables_initializer())

# 定义PyTorch模型
model = torch.nn.Sequential(
    torch.nn.Linear(784, 256),
    torch.nn.ReLU(),
    torch.nn.Linear(256, 10)
)

# 转换张量流的权重和偏置
with tf.variable_scope('', reuse=True):
    tf_vars = tf.trainable_variables()
    pytorch_vars = model.parameters()

    for tf_var, pytorch_var in zip(tf_vars, pytorch_vars):
        pytorch_var.data = torch.from_numpy(sess.run(tf_var))

# 定义训练循环
for epoch in range(num_epochs):
    # 执行训练操作
    sess.run(train_op, feed_dict={x: train_data, y: train_labels})

# 执行推理操作
output = sess.run(logits, feed_dict={x: test_data})

# 将PyTorch模型转换为推理模式
model.eval()
output = model(torch.from_numpy(test_data))