Update README.md
Browse filesadd quantized onnx model
README.md
CHANGED
@@ -10,12 +10,12 @@ widget:
|
|
10 |
DistilCamemBERT-NER
|
11 |
===================
|
12 |
|
13 |
-
We present DistilCamemBERT-NER which is [DistilCamemBERT](https://huggingface.co/cmarkea/distilcamembert-base) fine
|
14 |
|
15 |
Dataset
|
16 |
-------
|
17 |
|
18 |
-
The dataset used is [wikiner_fr](https://huggingface.co/datasets/Jean-Baptiste/wikiner_fr) which represents ~170k sentences
|
19 |
* PER: personality ;
|
20 |
* LOC: location ;
|
21 |
* ORG: organization ;
|
@@ -125,6 +125,22 @@ result
|
|
125 |
'end': 409}]
|
126 |
```
|
127 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
128 |
Citation
|
129 |
--------
|
130 |
```bibtex
|
|
|
10 |
DistilCamemBERT-NER
|
11 |
===================
|
12 |
|
13 |
+
We present DistilCamemBERT-NER, which is [DistilCamemBERT](https://huggingface.co/cmarkea/distilcamembert-base) fine-tuned for the NER (Named Entity Recognition) task for the French language. The work is inspired by [Jean-Baptiste/camembert-ner](https://huggingface.co/Jean-Baptiste/camembert-ner) based on the [CamemBERT](https://huggingface.co/camembert-base) model. The problem of the modelizations based on CamemBERT is at the scaling moment, for the production phase, for example. Indeed, inference cost can be a technological issue. To counteract this effect, we propose this modelization which **divides the inference time by two** with the same consumption power thanks to [DistilCamemBERT](https://huggingface.co/cmarkea/distilcamembert-base).
|
14 |
|
15 |
Dataset
|
16 |
-------
|
17 |
|
18 |
+
The dataset used is [wikiner_fr](https://huggingface.co/datasets/Jean-Baptiste/wikiner_fr), which represents ~170k sentences labeled in 5 categories :
|
19 |
* PER: personality ;
|
20 |
* LOC: location ;
|
21 |
* ORG: organization ;
|
|
|
125 |
'end': 409}]
|
126 |
```
|
127 |
|
128 |
+
### Optimum + ONNX
|
129 |
+
```python
|
130 |
+
from optimum.onnxruntime import ORTModelForTokenClassification
|
131 |
+
from transformers import AutoTokenizer, pipeline
|
132 |
+
|
133 |
+
HUB_MODEL = "cmarkea/distilcamembert-base-nli"
|
134 |
+
tokenizer = AutoTokenizer.from_pretrained(HUB_MODEL)
|
135 |
+
model = ORTModelForTokenClassification.from_pretrained(HUB_MODEL)
|
136 |
+
onnx_qa = pipeline("token-classification", model=model, tokenizer=tokenizer)
|
137 |
+
|
138 |
+
# Quantized onnx model
|
139 |
+
quantized_model = ORTModelForTokenClassification.from_pretrained(
|
140 |
+
HUB_MODEL, file_name="model_quantized.onnx"
|
141 |
+
)
|
142 |
+
```
|
143 |
+
|
144 |
Citation
|
145 |
--------
|
146 |
```bibtex
|