/
roberta-base-cola_houlsby.yaml
80 lines (63 loc) 路 2.56 KB
/
roberta-base-cola_houlsby.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
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
# Adapter-Hub adapter entry
# Defines a single adapter entry in Adapter-Hub
# --------------------
# The name of the author(s) of this adapter.
author: "Andreas R眉ckl茅"
# A bibtex citation of the work related to this adapter.
citation: |
@article{pfeiffer2020AdapterHub,
title={AdapterHub: A Framework for Adapting Transformers},
author={Jonas Pfeiffer,
Andreas R\"uckl\'{e},
Clifton Poth,
Aishwarya Kamath,
Ivan Vuli\'{c},
Sebastian Ruder,
Kyunghyun Cho,
Iryna Gurevych},
journal={ArXiv},
year={2020}
}
# The string identifier of the adapter architecture (must be available in architecture).
# Describes the adapter architecture used by this adapter
config: # TODO: REQUIRED
# The name of the adapter config used by this adapter (a short name available in the `architectures` folder).
# Example: pfeiffer
using: houlsby
# The version to be downloaded if no version is explicitly stated.
default_version: "1"
# A short description of this adapter.
description: |
Adapter (with head) trained using the `run_glue.py` script with an extension that retains the best checkpoint (out of 30 epochs).
# A contact email of the author(s).
email: "rueckle@ukp.informatik.tu-darmstadt.de"
# A list of different versions of this adapter available for download.
files:
- sha256: ab5d274ad64208676f3e313cd40297f9fb10b9518f3a34a72e73fcdc85c39768
sha1: "c568c8523945de5392aaee9145768cd90dd61472"
# Download URL pointing to a zip folder containing the adapter module.
url: "https://public.ukp.informatik.tu-darmstadt.de/AdapterHub/text_task/cola/roberta-base/houlsby/roberta-base_cola_houlsby.zip"
version: "1"
description: "Achieves 60.07 Matthew's corr on the CoLA devset"
# A GitHub handle associated with the author(s).
github: "arueckle"
# The hidden size of the model
hidden_size: 768
# The string identifier of the pre-trained model (by which it is identified at Huggingface).
# Example: bert-base-uncased
model_name: "roberta-base"
# The model type.
# Example: bert
model_type: roberta
# The string identifier of the subtask this adapter belongs to.
subtask: cola
# The string identifier of the task this adapter belongs to.
task: lingaccept
# A Twitter handle associated with the author(s).
twitter: "arueckle"
# The type of adapter (one of the options available in `adapter_type`.
type: "text_task"
# A URL providing more information on this adapter/ the authors/ the organization.
url: "http://rueckle.net"
model_class: RobertaModelWithHeads
prediction_head: true