Skip to content
This repository was archived by the owner on Aug 26, 2022. It is now read-only.

Commit 3bb2caa

Browse files
inesines
committed
Merge branch 'develop'
2 parents eca7b56 + 0e637ba commit 3bb2caa

File tree

19 files changed

+699
-723
lines changed

19 files changed

+699
-723
lines changed

‎README.md‎

Lines changed: 12 additions & 310 deletions
Original file line numberDiff line numberDiff line change
@@ -2,313 +2,15 @@
22

33
# spaCy REST services
44

5-
This repository provides REST microservices for Explosion AI's [interactive demos](https://demos.explosion.ai) and visualisers. All requests and responses are JSON-encoded as `text/string`, so all requests require the header `Content-Type: text/string`.
6-
7-
## [displaCy server](displacy)
8-
9-
A simple [Falcon](https://falconframework.org/) app for exposing a spaCy dependency parser and spaCy named entity recognition model as a REST microservice, formatted for the [displaCy.js](https://github.com/explosion/displacy) and [displaCy ENT](https://github.com/explosion/displacy-ent) visualiser. For more info on the rendering on the front-end that consumes the data produced by this service, see [this blog post](https://explosion.ai/blog/displacy-js-nlp-visualizer).
10-
11-
The service exposes two endpoints that accept POST requests, and two endpoints that accept GET requests to describe the available models and schemas.
12-
13-
---
14-
15-
### `POST` `/dep/`
16-
17-
Example request:
18-
19-
```json
20-
{
21-
"text": "They ate the pizza with anchovies",
22-
"model":"en",
23-
"collapse_punctuation": 0,
24-
"collapse_phrases": 1
25-
}
26-
```
27-
28-
| Name | Type | Description |
29-
| --- | --- | --- |
30-
| `text` | string | text to be parsed |
31-
| `model` | string | identifier string for a model installed on the server |
32-
| `collapse_punctuation` | boolean | Merge punctuation onto the preceding token? |
33-
| `collapse_phrases` | boolean | Merge noun chunks and named entities into single tokens? |
34-
35-
36-
Example request using the Python [Requests library](http://docs.python-requests.org/en/master/):
37-
38-
```python
39-
import json
40-
import requests
41-
42-
url = "http://localhost:8000/dep"
43-
message_text = "They ate the pizza with anchovies"
44-
headers = {'content-type': 'application/json'}
45-
d = {'text': message_text, 'model': 'en'}
46-
47-
response = requests.post(url, data=json.dumps(d), headers=headers)
48-
r = response.json()
49-
```
50-
51-
Example response:
52-
53-
```json
54-
{
55-
"arcs": [
56-
{ "dir": "left", "start": 0, "end": 1, "label": "nsubj" },
57-
{ "dir": "right", "start": 1, "end": 2, "label": "dobj" },
58-
{ "dir": "right", "start": 1, "end": 3, "label": "prep" },
59-
{ "dir": "right", "start": 3, "end": 4, "label": "pobj" },
60-
{ "dir": "left", "start": 2, "end": 3, "label": "prep" }
61-
],
62-
"words": [
63-
{ "tag": "PRP", "text": "They" },
64-
{ "tag": "VBD", "text": "ate" },
65-
{ "tag": "NN", "text": "the pizza" },
66-
{ "tag": "IN", "text": "with" },
67-
{ "tag": "NNS", "text": "anchovies" }
68-
]
69-
}
70-
```
71-
72-
| Name | Type | Description |
73-
| --- | --- | --- |
74-
| `arcs` | array | data to generate the arrows |
75-
| `dir` | string | direction of arrow (`"left"` or `"right"`) |
76-
| `start` | integer | offset of word the arrow starts **on** |
77-
| `end` | integer | offset of word the arrow ends **on** |
78-
| `label` | string | dependency label |
79-
| `words` | array | data to generate the words |
80-
| `tag` | string | part-of-speech tag |
81-
| `text` | string | token |
82-
83-
---
84-
85-
Curl command:
86-
87-
```
88-
curl -s localhost:8000/dep -d '{"text":"Pastafarians are smarter than people with Coca Cola bottles.", "model":"en"}'
89-
```
90-
91-
```json
92-
{
93-
"arcs": [
94-
{
95-
"dir": "left",
96-
"end": 1,
97-
"label": "nsubj",
98-
"start": 0
99-
},
100-
{
101-
"dir": "right",
102-
"end": 2,
103-
"label": "acomp",
104-
"start": 1
105-
},
106-
{
107-
"dir": "right",
108-
"end": 3,
109-
"label": "prep",
110-
"start": 2
111-
},
112-
{
113-
"dir": "right",
114-
"end": 4,
115-
"label": "pobj",
116-
"start": 3
117-
},
118-
{
119-
"dir": "right",
120-
"end": 5,
121-
"label": "prep",
122-
"start": 4
123-
},
124-
{
125-
"dir": "right",
126-
"end": 6,
127-
"label": "pobj",
128-
"start": 5
129-
}
130-
],
131-
"words": [
132-
{
133-
"tag": "NNPS",
134-
"text": "Pastafarians"
135-
},
136-
{
137-
"tag": "VBP",
138-
"text": "are"
139-
},
140-
{
141-
"tag": "JJR",
142-
"text": "smarter"
143-
},
144-
{
145-
"tag": "IN",
146-
"text": "than"
147-
},
148-
{
149-
"tag": "NNS",
150-
"text": "people"
151-
},
152-
{
153-
"tag": "IN",
154-
"text": "with"
155-
},
156-
{
157-
"tag": "NNS",
158-
"text": "Coca Cola bottles."
159-
}
160-
]
161-
}
162-
```
163-
164-
165-
### `POST` `/ent/`
166-
167-
Example request:
168-
169-
```json
170-
{
171-
"text": "When Sebastian Thrun started working on self-driving cars at Google in 2007, few people outside of the company took him seriously.",
172-
"model": "en"
173-
}
174-
```
175-
176-
| Name | Type | Description |
177-
| --- | --- | --- |
178-
| `text` | string | text to be parsed |
179-
| `model` | string | identifier string for a model installed on the server |
180-
181-
Example request using the Python [Requests library](http://docs.python-requests.org/en/master/):
182-
183-
```python
184-
import json
185-
import requests
186-
187-
url = "http://localhost:8000/ent"
188-
message_text = "When Sebastian Thrun started working on self-driving cars at Google in 2007, few people outside of the company took him seriously."
189-
headers = {'content-type': 'application/json'}
190-
d = {'text': message_text, 'model': 'en'}
191-
192-
response = requests.post(url, data=json.dumps(d), headers=headers)
193-
r = response.json()
194-
```
195-
196-
Example response:
197-
198-
```json
199-
[
200-
{ "end": 20, "start": 5, "type": "PERSON" },
201-
{ "end": 67, "start": 61, "type": "ORG" },
202-
{ "end": 75, "start": 71, "type": "DATE" }
203-
]
204-
```
205-
206-
| Name | Type | Description |
207-
| --- | --- | --- |
208-
| `end` | integer | character offset the entity ends **after** |
209-
| `start` | integer | character offset the entity starts **on** |
210-
| `type` | string | entity type |
211-
212-
213-
214-
```
215-
curl -s localhost:8000/ent -d '{"text":"Pastafarians are smarter than people with Coca Cola bottles.", "model":"en"}'
216-
```
217-
218-
```json
219-
[
220-
{
221-
"end": 12,
222-
"start": 0,
223-
"type": "NORP"
224-
},
225-
{
226-
"end": 51,
227-
"start": 42,
228-
"type": "ORG"
229-
}
230-
]
231-
```
232-
233-
234-
---
235-
236-
### `GET` `/models`
237-
238-
List the names of models installed on the server.
239-
240-
Example request:
241-
242-
```
243-
GET /models
244-
```
245-
246-
Example response:
247-
248-
```json
249-
["en", "de"]
250-
```
251-
252-
---
253-
254-
### `GET` `/{model}/schema/`
255-
256-
Example request:
257-
258-
```
259-
GET /en/schema
260-
```
261-
262-
| Name | Type | Description |
263-
| --- | --- | --- |
264-
| `model` | string | identifier string for a model installed on the server |
265-
266-
Example response:
267-
268-
```json
269-
{
270-
"dep_types": ["ROOT", "nsubj"],
271-
"ent_types": ["PERSON", "LOC", "ORG"],
272-
"pos_types": ["NN", "VBZ", "SP"]
273-
}
274-
```
275-
276-
---
277-
278-
## [sense2vec server](sense2vec)
279-
280-
A simple [Falcon](https://falconframework.org/) app for exposing a sense2vec model as a REST microservice, as used in the [sense2vec demo](https://github.com/explosion/sense2vec-demo)
281-
282-
The service exposes a single endpoint over GET.
283-
284-
---
285-
286-
### `GET` `/{word|POS}`
287-
288-
Example query:
289-
290-
```
291-
GET /natural_language_processing%7CNOUN
292-
```
293-
294-
Example response:
295-
296-
```json
297-
[
298-
{
299-
"score": 0.1,
300-
"key": "computational_linguistics|NOUN",
301-
"text": "computational linguistics",
302-
"count": 20,
303-
"head": "linguistics"
304-
}
305-
]
306-
```
307-
308-
| Name | Type | Description |
309-
| --- | --- | --- |
310-
| `score` | float | similarity to query |
311-
| `key` | string | identifier string |
312-
| `text` | string | human-readable token |
313-
| `count` | integer | absolute frequency in training corpus |
314-
| `head` | string | head word in text |
5+
This repository includes REST microservices for various spaCy-related tasks. The
6+
services power our [interactive demos](https://explosion.ai/demos) and can be
7+
used as examples of exposing spaCy's capabilities as a microservice. All APIs
8+
are built with [`hug`](https://github.com/timothycrosley/hug) and
9+
**require Python 3**. The following services are available – for more details,
10+
see the API docs in the respective directories.
11+
12+
| Service | Description | Example |
13+
| --- | --- | -- |
14+
| [`displacy`](displacy) | Serve a spaCy model and extract dependencies and named entities. | [🖼](https://explosion.ai/demos/displacy), [🖼](https://explosion.ai/demos/displacy-ent)|
15+
| [`sense2vec`](sense2vec) | Serve a [sense2vec](https://github.com/explosion/sense2vec) model with automatic sense detection. | [🖼](https://explosion.ai/demos/sense2vec)
16+
| [`matcher`](matcher) | Run a match pattern over a text and return the matches and tokens as JSON. | [🖼](https://explosion.ai/demos/matcher) |

0 commit comments

Comments
 (0)