phlippseitz commited on
Commit
a2dd466
1 Parent(s): 8091efe

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +60 -0
README.md ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - onnx
4
+ - question-answering
5
+ - roberta
6
+ - adapter-transformers
7
+ datasets:
8
+ - drop
9
+ language:
10
+ - en
11
+ ---
12
+
13
+ # ONNX export of Adapter `AdapterHub/roberta-base-pf-drop` for roberta-base
14
+ ## Conversion of [AdapterHub/roberta-base-pf-drop](https://huggingface.co/AdapterHub/roberta-base-pf-drop) for UKP SQuARE
15
+
16
+
17
+ ## Usage
18
+ ```python
19
+ onnx_path = hf_hub_download(repo_id='UKP-SQuARE/roberta-base-pf-drop-onnx', filename='model.onnx') # or model_quant.onnx for quantization
20
+ onnx_model = InferenceSession(onnx_path, providers=['CPUExecutionProvider'])
21
+
22
+ context = 'ONNX is an open format to represent models. The benefits of using ONNX include interoperability of frameworks and hardware optimization.'
23
+ question = 'What are advantages of ONNX?'
24
+ tokenizer = AutoTokenizer.from_pretrained('roberta-base')
25
+
26
+ inputs = tokenizer(question, context, padding=True, truncation=True, return_tensors='np')
27
+ inputs_int64 = {key: np.array(inputs[key], dtype=np.int64) for key in inputs}
28
+ outputs = onnx_model.run(input_feed=dict(inputs_int64), output_names=None)
29
+ ```
30
+
31
+ ## Architecture & Training
32
+
33
+ The training code for this adapter is available at https://github.com/adapter-hub/efficient-task-transfer.
34
+ In particular, training configurations for all tasks can be found [here](https://github.com/adapter-hub/efficient-task-transfer/tree/master/run_configs).
35
+
36
+
37
+ ## Evaluation results
38
+
39
+ Refer to [the paper](https://arxiv.org/pdf/2104.08247) for more information on results.
40
+
41
+ ## Citation
42
+
43
+ If you use this adapter, please cite our paper ["What to Pre-Train on? Efficient Intermediate Task Selection"](https://arxiv.org/pdf/2104.08247):
44
+
45
+ ```bibtex
46
+ @inproceedings{poth-etal-2021-pre,
47
+ title = "{W}hat to Pre-Train on? {E}fficient Intermediate Task Selection",
48
+ author = {Poth, Clifton and
49
+ Pfeiffer, Jonas and
50
+ R{"u}ckl{'e}, Andreas and
51
+ Gurevych, Iryna},
52
+ booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing",
53
+ month = nov,
54
+ year = "2021",
55
+ address = "Online and Punta Cana, Dominican Republic",
56
+ publisher = "Association for Computational Linguistics",
57
+ url = "https://aclanthology.org/2021.emnlp-main.827",
58
+ pages = "10585--10605",
59
+ }
60
+ ```