$ cd oneie
$ python app.py --port 61802 --tmp_dir <path/to/the/dir/for/tmp/files>
To call the API, send a POST request with application/json as mimetype and a
JSON format string as its data with the following fields:
{
// Required field
"data": [
{
"doc_id": "xxxx01",
"doc_tokens": [
[['a', 0, 1], ['sentence', 2, 10]],
[['sentence', 11, 19], ['2', 20, 21]]
]
},
...
],
// Optional fields
"lang": "en",
"batch_size": 10,
"max_length": 128,
"beam_size": 10,
"device": 0
}
The value of the data field in this JSON string is a list. Each object in the
list is a document with its document ID doc_id and tokens doc_tokens. Each
document should have a unique doc_id. The value of the doc_tokens field is
a list of lists. Each list contains tokens of a sentence in the document, where
each token is a triplet of (token_text, start_char_offset, end_char_offset).
The API returns a JSON object as follows:
{
'oneie': {
# OneIE's JSON format output files
'json': {
'doc_id_1': ...,
'doc_id_2': ...,
...
},
# Cold-start format output files
'cs': {
'entity': ...,
'relation': ...,
'event': ...
},
# TAB format output files
'tab': {
'nam': ...,
'nom': ...,
'pro': ...,
'nam+nom+pro': ...
},
# BIO format output files
'bio': {
'nam': ...,
'nom': ...,
'pro': ...,
'nam+nom+pro': ...
},
# Fine-grained entity typing output file
'cfet': ...
}
}
The value of each field is the text content of the corresponding file.
Content type
Image
Digest
Size
10.9 GB
Last updated
almost 6 years ago
docker pull limteng/oneie_kairos