Add handler.py to support inference endpoints
Browse files- handler.py +33 -0
handler.py
ADDED
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from typing import Any, Dict, List
|
2 |
+
|
3 |
+
from span_marker import SpanMarkerModel
|
4 |
+
|
5 |
+
|
6 |
+
class EndpointHandler:
|
7 |
+
def __init__(self, model_id: str) -> None:
|
8 |
+
self.model = SpanMarkerModel.from_pretrained(model_id)
|
9 |
+
# Try to place it on CUDA, do nothing if it fails
|
10 |
+
self.model.try_cuda()
|
11 |
+
|
12 |
+
def __call__(self, data: Dict[str, Any]) -> List[Dict[str, Any]]:
|
13 |
+
"""
|
14 |
+
Args:
|
15 |
+
data (Dict[str, Any]):
|
16 |
+
a dictionary with the "inputs" key corresponding to a string containing some text
|
17 |
+
Return:
|
18 |
+
A List[Dict[str, Any]]:. The object returned should be like [{"entity_group": "XXX", "word": "some word", "start": 3, "end": 6, "score": 0.82}] containing :
|
19 |
+
- "entity_group": A string representing what the entity is.
|
20 |
+
- "word": A rubstring of the original string that was detected as an entity.
|
21 |
+
- "start": the offset within `input` leading to `answer`. context[start:stop] == word
|
22 |
+
- "end": the ending offset within `input` leading to `answer`. context[start:stop] === word
|
23 |
+
- "score": A score between 0 and 1 describing how confident the model is for this entity.
|
24 |
+
"""
|
25 |
+
return [
|
26 |
+
{
|
27 |
+
"entity_group": entity["label"],
|
28 |
+
"word": entity["span"],
|
29 |
+
"start": entity["char_start_index"],
|
30 |
+
"end": entity["char_end_index"],
|
31 |
+
"score": entity["score"],
|
32 |
+
}
|
33 |
+
for entity in self.model.predict(data["inputs"])
|