diff --git a/guides/custom_train_step_in_torch.py b/guides/custom_train_step_in_torch.py
index 07a89cfed4..b493411ab5 100644
--- a/guides/custom_train_step_in_torch.py
+++ b/guides/custom_train_step_in_torch.py
@@ -2,7 +2,7 @@
Title: Customizing what happens in `fit()` with PyTorch
Author: [fchollet](https://twitter.com/fchollet)
Date created: 2023/06/27
-Last modified: 2023/06/27
+Last modified: 2024/08/01
Description: Overriding the training step of the Model class with PyTorch.
Accelerator: GPU
"""
@@ -397,7 +397,7 @@ def compile(self, d_optimizer, g_optimizer, loss_fn):
def train_step(self, real_images):
device = "cuda" if torch.cuda.is_available() else "cpu"
- if isinstance(real_images, tuple):
+ if isinstance(real_images, tuple) or isinstance(real_images, list):
real_images = real_images[0]
# Sample random points in the latent space
batch_size = real_images.shape[0]
diff --git a/guides/ipynb/custom_train_step_in_torch.ipynb b/guides/ipynb/custom_train_step_in_torch.ipynb
index 0ae7a8b9f0..9f3d3576d1 100644
--- a/guides/ipynb/custom_train_step_in_torch.ipynb
+++ b/guides/ipynb/custom_train_step_in_torch.ipynb
@@ -10,7 +10,7 @@
"\n",
"**Author:** [fchollet](https://twitter.com/fchollet)
\n",
"**Date created:** 2023/06/27
\n",
- "**Last modified:** 2023/06/27
\n",
+ "**Last modified:** 2024/08/01
\n",
"**Description:** Overriding the training step of the Model class with PyTorch."
]
},
@@ -50,17 +50,6 @@
"Let's see how that works."
]
},
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {
- "colab_type": "code"
- },
- "outputs": [],
- "source": [
- "!pip install keras --upgrade --quiet"
- ]
- },
{
"cell_type": "markdown",
"metadata": {
@@ -282,7 +271,7 @@
"outputs = keras.layers.Dense(1)(inputs)\n",
"model = CustomModel(inputs, outputs)\n",
"\n",
- "# We don't passs a loss or metrics here.\n",
+ "# We don't pass a loss or metrics here.\n",
"model.compile(optimizer=\"adam\")\n",
"\n",
"# Just use `fit` as usual -- you can use callbacks, etc.\n",
@@ -531,7 +520,7 @@
"\n",
" def train_step(self, real_images):\n",
" device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
- " if isinstance(real_images, tuple):\n",
+ " if isinstance(real_images, tuple) or isinstance(real_images, list):\n",
" real_images = real_images[0]\n",
" # Sample random points in the latent space\n",
" batch_size = real_images.shape[0]\n",
diff --git a/guides/md/custom_train_step_in_torch.md b/guides/md/custom_train_step_in_torch.md
index c112a44be6..0b62d818f0 100644
--- a/guides/md/custom_train_step_in_torch.md
+++ b/guides/md/custom_train_step_in_torch.md
@@ -2,7 +2,7 @@
**Author:** [fchollet](https://twitter.com/fchollet)
**Date created:** 2023/06/27
-**Last modified:** 2023/06/27
+**Last modified:** 2024/08/01
**Description:** Overriding the training step of the Model class with PyTorch.
@@ -142,14 +142,14 @@ model.fit(x, y, epochs=3)
```
-Epoch 1/3
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 551us/step - mae: 0.6533 - loss: 0.6036
+Epoch 1/3
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 2ms/step - mae: 0.3410 - loss: 0.1772
Epoch 2/3
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 522us/step - mae: 0.4013 - loss: 0.2522
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 2ms/step - mae: 0.3336 - loss: 0.1695
Epoch 3/3
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 516us/step - mae: 0.3813 - loss: 0.2256
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 3ms/step - mae: 0.3170 - loss: 0.1511
-
+
```
@@ -238,17 +238,17 @@ model.fit(x, y, epochs=5)
```
Epoch 1/5
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 461us/step - loss: 0.2470 - mae: 0.3953
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 2ms/step - loss: 0.6173 - mae: 0.6607
Epoch 2/5
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 456us/step - loss: 0.2386 - mae: 0.3910
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 2ms/step - loss: 0.2340 - mae: 0.3883
Epoch 3/5
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 456us/step - loss: 0.2359 - mae: 0.3901
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 2ms/step - loss: 0.1922 - mae: 0.3517
Epoch 4/5
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 480us/step - loss: 0.2013 - mae: 0.3572
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 2ms/step - loss: 0.1802 - mae: 0.3411
Epoch 5/5
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 463us/step - loss: 0.1903 - mae: 0.3480
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 3ms/step - loss: 0.1862 - mae: 0.3505
-
+
```
@@ -328,13 +328,13 @@ model.fit(x, y, sample_weight=sw, epochs=3)
```
Epoch 1/3
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 499us/step - mae: 1.4332 - loss: 1.0769
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 2ms/step - mae: 0.3216 - loss: 0.0827
Epoch 2/3
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 520us/step - mae: 0.9250 - loss: 0.5614
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 2ms/step - mae: 0.3156 - loss: 0.0803
Epoch 3/3
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 502us/step - mae: 0.6069 - loss: 0.2653
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 2ms/step - mae: 0.3085 - loss: 0.0760
-
+
```
@@ -378,11 +378,23 @@ y = np.random.random((1000, 1))
model.evaluate(x, y)
```
+
+ 1/32 [37m━━━━━━━━━━━━━━━━━━━━ 0s 2ms/step - mae: 0.8706 - loss: 0.9344
+
```
- 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 325us/step - mae: 0.4427 - loss: 0.2993
+
+```
+
+ 32/32 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - mae: 0.8959 - loss: 0.9952
+
+
-[0.2726495862007141, 0.42286917567253113]
+
+
+
+```
+[1.0077838897705078, 0.8984771370887756]
```
@@ -463,7 +475,7 @@ class GAN(keras.Model):
def train_step(self, real_images):
device = "cuda" if torch.cuda.is_available() else "cpu"
- if isinstance(real_images, tuple):
+ if isinstance(real_images, tuple) or isinstance(real_images, list):
real_images = real_images[0]
# Sample random points in the latent space
batch_size = real_images.shape[0]
@@ -556,9 +568,8 @@ gan.fit(dataloader, epochs=1)
```
- 1094/1094 ━━━━━━━━━━━━━━━━━━━━ 1582s 1s/step - d_loss: 0.3581 - g_loss: 2.0571
-
-
+ 1094/1094 ━━━━━━━━━━━━━━━━━━━━ 394s 360ms/step - d_loss: 0.2436 - g_loss: 4.7259
+
```