thinc crystal_ball A refreshing functional take on deep learning, compatible with your favorite libraries

See also

• https://github.com/explosion/thinc

• https://fediverse.org/explosion_ai

• https://fediverse.org/honnibal

• https://thinc.ai/

• https://thinc.ai/docs/usage-type-checking

Annonces

See also

• https://fediverse.org/PyTorch/status/1222281229095096324?s=20

• https://fediverse.org/deliprao/status/1222301223690260480?s=20

• https://fediverse.org/honnibal/status/1222272108547407873?s=20

Features

🚀 Type checking

See also

• https://thinc.ai/docs/usage-type-checking

Develop faster and catch bugs sooner with sophisticated type checking.

Trying to pass a 1-dimensional array into a model that expects 2 dimensions ? That’s a type error. Your editor can pick it up as the code leaves your fingers.

🐍 Awesome config

See also

• https://thinc.ai/docs/usage-config

Configuration is a major pain for ML.

Thinc lets you describe trees of objects, with references to your own functions, so you can stop passing around blobs of settings. It’s simple, clean, and it works for both research and production.

🔮 Use any framework

See also

• https://thinc.ai/docs/usage-frameworks

Switch between PyTorch, TensorFlow and MXNet models without changing your application, or even create mutant hybrids using zero-copy array interchange.