lazytorch 0.1.0

The easiest ML library you'll ever use.

lazytorch

The easiest ML library you'll ever use. Currently in early beta. Do not use for production yet.

What is LazyTorch?

LazyTorch is a ML library that lets you make neural networks and models with ease. Just like the name suggests, LazyTorch is for lazy people like me.

Code Examples

Brain - General classification and neural networks.

model = Brain(input={"text": ""}, output={"is_positive": True})

ds = Dataset(
    {"text": [
        "i love this",
        "this is great",
        "absolute perfection",
        "i hate this",
        "this is terrible",
        "this is awful"
    ]},
    {"is_positive": [True, True, True, False, False, False]}
)

model.teach(ds)

print(model.infer({"text": "i love this"})) # outputs: {'is_positive': True}
print(model.infer({"text": "this is terrible"})) # outputs: {'is_positive': False}

AutoregressiveBrain - Brain but for autoregressive text generation and word prediction.

model = AutoregressiveBrain(context_size=2)
model.teach("the capital of russia is moscow. the capital of britan is london. the capital of france is paris. the capital of greece is athens.")
print(model.generate("the capital of france is", max_words=1)) # output: paris

Dataset - Datasets.

ds = Dataset({"text": ["hello", "world"]}, {"is_positive": [True, True]})
model = Brain(input={"text": ""}, output={"is_positive": True})
model.teach(ds)

BrainTrust - Sample results from multiple Brain models to get a general consensus.

ensemble = BrainTrust([model_a, model_b])
print(ensemble.infer({"text": "spam message"})) # output: {'is_spam': True}

BrainScanner - Evaluate Brain models on datasets.

accuracy = BrainScanner.evaluate(model, ds)
print(accuracy) # output: 1.0 (100%)

These are all of the functions available in the beta of LazyTorch. These are not final.