MODELS.md

# Model index

Every artifact here, what it is, and how to load it.

## `00_ai_universe_overview.png`

- **Type:** figure from the book
- **Size:** 0.51 MB

```python
# image
```

## `01_expert_system.png`

- **Type:** figure from the book
- **Size:** 0.04 MB

```python
# image
```

## `01_planning_search.png`

- **Type:** figure from the book
- **Size:** 0.03 MB

```python
# image
```

## `01x_fuzzy_logic.png`

- **Type:** figure from the book
- **Size:** 0.12 MB

```python
# image
```

## `02_ensemble.png`

- **Type:** figure from the book
- **Size:** 0.04 MB

```python
# image
```

## `02_kmeans_pca.png`

- **Type:** figure from the book
- **Size:** 0.15 MB

```python
# image
```

## `02_ml_classifiers.png`

- **Type:** figure from the book
- **Size:** 0.17 MB

```python
# image
```

## `02_pca_tsne.png`

- **Type:** figure from the book
- **Size:** 0.05 MB

```python
# image
```

## `02_regression.png`

- **Type:** figure from the book
- **Size:** 0.09 MB

```python
# image
```

## `02x_feature_engineering.png`

- **Type:** figure from the book
- **Size:** 0.16 MB

```python
# image
```

## `02x_semi_supervised.png`

- **Type:** figure from the book
- **Size:** 0.05 MB

```python
# image
```

## `03_activation_functions.png`

- **Type:** figure from the book
- **Size:** 0.13 MB

```python
# image
```

## `03_cnn_filters.png`

- **Type:** figure from the book
- **Size:** 0.02 MB

```python
# image
```

## `03_lstm_rnn.png`

- **Type:** figure from the book
- **Size:** 0.21 MB

```python
# image
```

## `03_mlp_confusion.png`

- **Type:** figure from the book
- **Size:** 0.06 MB

```python
# image
```

## `03_mlp_mnist.png`

- **Type:** figure from the book
- **Size:** 0.06 MB

```python
# image
```

## `03_perceptron.png`

- **Type:** figure from the book
- **Size:** 0.07 MB

```python
# image
```

## `03_som.png`

- **Type:** figure from the book
- **Size:** 0.07 MB

```python
# image
```

## `03x_backpropagation.png`

- **Type:** figure from the book
- **Size:** 0.10 MB

```python
# image
```

## `04_attention.png`

- **Type:** figure from the book
- **Size:** 0.13 MB

```python
# image
```

## `04_cifar10_samples.png`

- **Type:** figure from the book
- **Size:** 0.11 MB

```python
# image
```

## `04_dropout.png`

- **Type:** figure from the book
- **Size:** 0.10 MB

```python
# image
```

## `04_gan_generated.png`

- **Type:** figure from the book
- **Size:** 0.05 MB

```python
# image
```

## `04_gan_loss.png`

- **Type:** figure from the book
- **Size:** 0.04 MB

```python
# image
```

## `04_qlearning.png`

- **Type:** figure from the book
- **Size:** 0.09 MB

```python
# image
```

## `04_transfer_learning.png`

- **Type:** figure from the book
- **Size:** 0.06 MB

```python
# image
```

## `04x_capsnet.png`

- **Type:** figure from the book
- **Size:** 0.06 MB

```python
# image
```

## `04x_dbn_filters.png`

- **Type:** figure from the book
- **Size:** 0.17 MB

```python
# image
```

## `05_chatbot_similarity.png`

- **Type:** figure from the book
- **Size:** 0.14 MB

```python
# image
```

## `05_ngram_lm.png`

- **Type:** figure from the book
- **Size:** 0.05 MB

```python
# image
```

## `05_transformer.png`

- **Type:** figure from the book
- **Size:** 0.06 MB

```python
# image
```

## `07_mamba_loss.png`

- **Type:** figure from the book
- **Size:** 0.03 MB

```python
# image
```

## `07_stretch_goals_summary.png`

- **Type:** figure from the book
- **Size:** 0.24 MB

```python
# image
```

## `07_summary_table.png`

- **Type:** figure from the book
- **Size:** 0.26 MB

```python
# image
```

## `models/cnn_mnist.keras`

- **Type:** Keras full model - loads standalone
- **Size:** 1.18 MB

```python
from tensorflow import keras
model = keras.models.load_model('models/cnn_mnist.keras')
```

## `models/dnn_cifar.keras`

- **Type:** Keras full model - loads standalone
- **Size:** 22.17 MB

```python
from tensorflow import keras
model = keras.models.load_model('models/dnn_cifar.keras')
```

## `models/dropout_mnist.keras`

- **Type:** Keras full model - loads standalone
- **Size:** 2.85 MB

```python
from tensorflow import keras
model = keras.models.load_model('models/dropout_mnist.keras')
```

## `models/ensemble_voting.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 20.46 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/ensemble_voting.skops')   # review this list before trusting
model = load('models/ensemble_voting.skops', trusted=u)
```

## `models/expert_rules.json`

- **Type:** configuration / symbolic state
- **Size:** 0.00 MB

```python
import json
cfg = json.load(open('models/expert_rules.json'))
```

## `models/features_randomforest.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 3.88 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/features_randomforest.skops')   # review this list before trusting
model = load('models/features_randomforest.skops', trusted=u)
```

## `models/fuzzy_controller.json`

- **Type:** configuration / symbolic state
- **Size:** 0.00 MB

```python
import json
cfg = json.load(open('models/fuzzy_controller.json'))
```

## `models/gan_discriminator.keras`

- **Type:** Keras full model - loads standalone
- **Size:** 6.43 MB

```python
from tensorflow import keras
model = keras.models.load_model('models/gan_discriminator.keras')
```

## `models/hybrid_qnn.npz`

- **Type:** NumPy archive - trained circuit parameters
- **Size:** 0.00 MB

```python
import numpy as np
p = np.load('models/hybrid_qnn.npz')
print(p.files)
```

## `models/iris_decisiontree.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 0.01 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/iris_decisiontree.skops')   # review this list before trusting
model = load('models/iris_decisiontree.skops', trusted=u)
```

## `models/iris_knn.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 0.02 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/iris_knn.skops')   # review this list before trusting
model = load('models/iris_knn.skops', trusted=u)
```

## `models/iris_logreg.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 0.01 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/iris_logreg.skops')   # review this list before trusting
model = load('models/iris_logreg.skops', trusted=u)
```

## `models/iris_naivebayes.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 0.01 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/iris_naivebayes.skops')   # review this list before trusting
model = load('models/iris_naivebayes.skops', trusted=u)
```

## `models/iris_randomforest.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 1.82 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/iris_randomforest.skops')   # review this list before trusting
model = load('models/iris_randomforest.skops', trusted=u)
```

## `models/iris_scaler.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 0.01 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/iris_scaler.skops')   # review this list before trusting
model = load('models/iris_scaler.skops', trusted=u)
```

## `models/iris_svm.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 0.02 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/iris_svm.skops')   # review this list before trusting
model = load('models/iris_svm.skops', trusted=u)
```

## `models/kmeans_iris.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 0.01 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/kmeans_iris.skops')   # review this list before trusting
model = load('models/kmeans_iris.skops', trusted=u)
```

## `models/lstm_sine.keras`

- **Type:** Keras full model - loads standalone
- **Size:** 0.26 MB

```python
from tensorflow import keras
model = keras.models.load_model('models/lstm_sine.keras')
```

## `models/manifest.json`

- **Type:** configuration / symbolic state
- **Size:** 0.00 MB

```python
import json
cfg = json.load(open('models/manifest.json'))
```

## `models/mlp_mnist.keras`

- **Type:** Keras full model - loads standalone
- **Size:** 2.85 MB

```python
from tensorflow import keras
model = keras.models.load_model('models/mlp_mnist.keras')
```

## `models/planner_astar.json`

- **Type:** configuration / symbolic state
- **Size:** 0.00 MB

```python
import json
cfg = json.load(open('models/planner_astar.json'))
```

## `models/qaoa_maxcut.npz`

- **Type:** NumPy archive - trained circuit parameters
- **Size:** 0.00 MB

```python
import numpy as np
p = np.load('models/qaoa_maxcut.npz')
print(p.files)
```

## `models/qgan_generator.npz`

- **Type:** NumPy archive - trained circuit parameters
- **Size:** 0.00 MB

```python
import numpy as np
p = np.load('models/qgan_generator.npz')
print(p.files)
```

## `models/qsvm_kernel.npz`

- **Type:** NumPy archive - trained circuit parameters
- **Size:** 0.00 MB

```python
import numpy as np
p = np.load('models/qsvm_kernel.npz')
print(p.files)
```

## `models/regression_poly3.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 0.02 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/regression_poly3.skops')   # review this list before trusting
model = load('models/regression_poly3.skops', trusted=u)
```

## `models/semisup_labelspreading.skops`

- **Type:** scikit-learn model (skops, safer than pickle)
- **Size:** 1.09 MB

```python
from skops.io import load, get_untrusted_types
u = get_untrusted_types(file='models/semisup_labelspreading.skops')   # review this list before trusting
model = load('models/semisup_labelspreading.skops', trusted=u)
```

## `models/transfer_cifar.keras`

- **Type:** Keras full model - loads standalone
- **Size:** 13.59 MB

```python
from tensorflow import keras
model = keras.models.load_model('models/transfer_cifar.keras')
```

## `models/vqc_iris.npz`

- **Type:** NumPy archive - trained circuit parameters
- **Size:** 0.00 MB

```python
import numpy as np
p = np.load('models/vqc_iris.npz')
print(p.files)
```

## `models/vqe_ising.npz`

- **Type:** NumPy archive - trained circuit parameters
- **Size:** 0.00 MB

```python
import numpy as np
p = np.load('models/vqe_ising.npz')
print(p.files)
```