Generate statistical information of the availability of tokens or compounds in texts.
git clone git@github.com:StephanGeorg/token-stats.git
cd token-stats
npm i
Download models from nnsplit repo.
> npm run cli -- <path/to/input.csv> </path/to/output.csv> -m </path/to/model.onnx> [-s] [-t] [-c]| Parameter | Default | Description |
|---|---|---|
| Input | Required | Path to input file |
| Output | Required | Path to output file |
| -m | Required | Path to the model file (onnx) |
| -s | 0 | Index of desired sentence in text |
| -t | Index of desired token in sentence, -1 for last or empty | |
| -c | Index of desired constituent in token, -1 for last or empty |
name,count
straße,126369
weg,32322
platz,8271
berg,8177
gasse,6808
ring,3577
feld,3516
kamp,3476
allee,2763
mühle,2576
brücke,2518
bach,2482
garten,2346
graben,2265
grund,1665
wiese,1661
damm,1557
pfad,1524
busch,1472