rthy

4 ms·

1. 1.▲
   TF.text: Text Processing in Tensorflow(medium.com)

   2 points by rthy 7y ago|0 comments
2. 2.▲
   by rthy 7y ago

   I have been experimenting with tokenization in Rust in https://github.com/rth/vtext , mostly relying on Unicode segmentation with a few custom rules for tokenization. The obtained performance was also around 10x faster
3. 3.▲
   by rthy 8y ago

   Alternatively, the toolz package ( https://toolz.readthedocs.io/en/latest/ ) is a nice way of getting some additional functional programming capabilities while using the standard CPython interpreter.
4. 4.▲
   by rthy 8y ago

   An example of compiling the CPython interpreter to WebAssembly can be found in Pyodide ( https://github.com/iodide-project/pyodide/ ).