/
roberta-base_qa_squad2_pfeiffer.yaml
59 lines (45 loc) 路 2.01 KB
/
roberta-base_qa_squad2_pfeiffer.yaml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
# Adapter-Hub adapter entry
# Defines a single adapter entry in Adapter-Hub
# --------------------
# The type of adapter (one of the options available in `adapter_type`.
type: text_task
# The string identifier of the task this adapter belongs to.
task: qa
# The string identifier of the subtask this adapter belongs to.
subtask: squad2
# The model type.
# Example: bert
model_type: roberta
# The string identifier of the pre-trained model (by which it is identified at Huggingface).
# Example: bert-base-uncased
model_name: roberta-base
# The name of the author(s) of this adapter.
author: Clifton Poth
# Describes the adapter architecture used by this adapter
config:
# The name of the adapter config used by this adapter (a short name available in the `architectures` folder).
# Example: pfeiffer
using: pfeiffer
default_version: '1'
# A list of different versions of this adapter available for download.
files:
- version: '1'
url: https://public.ukp.informatik.tu-darmstadt.de/AdapterHub/text_task/squad2/roberta-base/pfeiffer/roberta-base_qa_squad2_pfeiffer.zip
sha1: 613711b48b651dd10e8c1e69af0f40b48e0c6dfb
sha256: 83a3eb224c18e926d90346b21938c50c3fb3af60a14c7acb94242c28a14d582d
score: 80.35
# (optional) A short description of this adapter.
description: |
Adapter for roberta-base in Pfeiffer architecture trained on the SQuAD 2.0 dataset for 15 epochs with early stopping and a learning rate of 1e-4.
# (optional) A contact email of the author(s).
email: poth@ukp.informatik.tu-darmstadt.de
# (optional) A GitHub handle associated with the author(s).
github: calpt
# (optional) The name of the model class from which this adapter was extracted. This field is mainly intended for adapters with prediction heads.
# Example: BertModelWithHeads
model_class: RobertaModelWithHeads
# (optional) A Twitter handle associated with the author(s).
twitter: '@clifapt'
# (optional) A URL providing more information on this adapter/ the authors/ the organization.
url: https://www.informatik.tu-darmstadt.de/ukp
prediction_head: true