spaCy/website/docs/usage/facts-figures.md

60 lines
3.8 KiB
Markdown
Raw Normal View History

---
title: Facts & Figures
teaser: The hard numbers for spaCy and how it compares to other tools
next: /usage/spacy-101
menu:
- ['Feature Comparison', 'comparison']
- ['Benchmarks', 'benchmarks']
2020-09-12 15:05:10 +00:00
# TODO: - ['Citing spaCy', 'citation']
---
2020-09-12 15:05:10 +00:00
## Comparison {#comparison hidden="true"}
2020-09-12 15:05:10 +00:00
### When should I use spaCy? {#comparison-usage}
2020-09-12 15:05:10 +00:00
<!-- TODO: update -->
2020-09-12 15:05:10 +00:00
| Use Cases |
| ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| ✅ **I'm a beginner and just getting started with NLP.**<br />spaCy makes it easy to get started and comes with extensive documentation, including a beginner-friendly [101 guide](/usage/spacy-101) and a free interactive [online course](https://course.spacy.io). |
| ✅ **I want to build an end-to-end production application.** |
| ✅ **I want my application to be efficient on CPU.**<br />While spaCy lets you train modern NLP models that are best run on GPU, it also offers CPU-optimized pipelines, which may be less accurate but much cheaper to run. |
| ✅ **I want to try out different neural network architectures for NLP.** |
| ❌ **I want to build a language generation application.**<br />spaCy's focus is natural language _processing_ and extracting information from large volumes of text. While you can use it to help you re-write existing text, it doesn't include any specific functionality for language generation tasks. |
| ❌ **I want to research machine learning algorithms.** |
## Benchmarks {#benchmarks}
2020-09-12 15:05:10 +00:00
spaCy v3.0 introduces transformer-based pipelines that bring spaCy's accuracy
right up to **current state-of-the-art**. You can also use a CPU-optimized
pipeline, which is less accurate but much cheaper to run.
2020-09-12 15:05:10 +00:00
<!-- TODO: -->
2020-09-12 15:05:10 +00:00
> #### Evaluation details
>
2020-09-12 15:05:10 +00:00
> - **OntoNotes 5.0:** spaCy's English models are trained on this corpus, as
> it's several times larger than other English treebanks. However, most
> systems do not report accuracies on it.
> - **Penn Treebank:** The "classic" parsing evaluation for research. However,
> it's quite far removed from actual usage: it uses sentences with
> gold-standard segmentation and tokenization, from a pretty specific type of
> text (articles from a single newspaper, 1984-1989).
2020-09-12 15:05:10 +00:00
import Benchmarks from 'usage/\_benchmarks-models.md'
2020-09-12 15:05:10 +00:00
<Benchmarks />
2020-09-12 15:05:10 +00:00
<!-- TODO: update -->
2020-09-12 15:05:10 +00:00
<Project id="benchmarks/penn_treebank">
2020-09-12 15:05:10 +00:00
The easiest way to reproduce spaCy's benchmarks on the Penn Treebank is to clone
our project template.
2020-09-12 15:05:10 +00:00
</Project>
2020-09-12 15:05:10 +00:00
<!-- ## Citing spaCy {#citation}
2020-09-12 15:05:10 +00:00
<!-- TODO: update -->