前向传播、损失、反向传播与更新
一次训练迭代的四个步骤
前向传播、损失、反向传播与更新 是 CoddyKit 上的免费 Deep Learning Academy 课时。 这是第 1 节课，共 4 节。 你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。 这是 Deep Learning Academy 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 Deep Learning Academy 课程共包含 4 节课。
本课时的部分内容尚未翻译，以英文显示。
Four Moves, One Beat
Every training iteration is just four moves repeated: forward, loss, backward, step. Learn this rhythm and the whole training loop stops feeling mysterious. 🥁
Move One: Forward
The forward pass sends your input through the model so it produces a prediction. You just call the model on a batch of data and read the output.
pred = model(x)Logits, Not Answers Yet
The forward output is usually raw scores called logits, not clean labels. The loss function knows how to read these scores directly, so you leave them as is.
Move Two: Loss
Next you measure how wrong the prediction was. The loss compares your output to the true labels and boils the mistake down to one number.
loss = loss_fn(pred, y)Lower Loss, Better Model
A big loss means a poor prediction and a small one means a good prediction. Training has a single mission: drive this number steadily downward.
Move Three: Backward
Now you ask which way to nudge each weight. The backward pass uses autograd to compute every gradient and fill each parameter's .grad for you.
loss.backward()Gradients Are Directions
Each gradient tells you how the loss would change if you tweaked that one weight. They are the map the optimizer follows toward a better model.
Move Four: Step
Finally the optimizer applies those gradients, nudging every weight a little. This step is the moment the model actually learns from its mistake.
optimizer.step()Clear the Grads
Gradients pile up by default, so before the next backward you call zero_grad. Skip it and old gradients corrupt your next update.
optimizer.zero_grad()The Whole Move
Put the four moves together and you get one clean iteration. You will type this exact pattern in nearly every PyTorch project you build.
optimizer.zero_grad()
pred = model(x)
loss = loss_fn(pred, y)
loss.backward()
optimizer.step()Then Do It Again
Run these moves on batch after batch and the loss falls a little each time. Thousands of tiny updates are what turn random weights into real skill.
Quick Check
Which call actually computes the gradients for every weight?
Recap
One iteration is four moves: forward for a prediction, loss to measure error, backward for gradients, and step to update weights. Repeat to learn. 🎉
常见问题解答
「前向传播、损失、反向传播与更新」课时是免费的吗？
是的 — 「前向传播、损失、反向传播与更新」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 Deep Learning Academy 课程的其余内容，请升级到 CoddyKit PRO。 Deep Learning Academy 课程共包含 4 节课。
「前向传播、损失、反向传播与更新」这节课中我会学到什么？
一次训练迭代的四个步骤 你通过在浏览器中直接运行的动手代码来练习 Deep Learning Academy，全天候 AI 导师会在你学习这节课的过程中回答你的问题。
学习 Deep Learning Academy 需要有经验吗？
无需任何先前经验。CoddyKit 上的 Deep Learning Academy 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。 这是第 1 节课，共 4 节。
「前向传播、损失、反向传播与更新」课时需要多长时间？
大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。
我能在这节 Deep Learning Academy 课中编写并运行代码吗？
能。每节 Deep Learning Academy 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。