Add adapter bert-base-multilingual-cased-mi-wiki_houlsby version nd
Browse files- ._adapter_config.json +0 -0
- ._mi_houlsby_gelu_nd +0 -0
- README.md +65 -0
- adapter_config.json +41 -0
- pytorch_adapter.bin +3 -0
._adapter_config.json
ADDED
Binary file (276 Bytes). View file
|
|
._mi_houlsby_gelu_nd
ADDED
Binary file (310 Bytes). View file
|
|
README.md
ADDED
@@ -0,0 +1,65 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- adapterhub:mi/wiki
|
4 |
+
- bert
|
5 |
+
- adapter-transformers
|
6 |
+
language:
|
7 |
+
- mi
|
8 |
+
license: "apache-2.0"
|
9 |
+
---
|
10 |
+
|
11 |
+
# Adapter `bert-base-multilingual-cased-mi-wiki_houlsby` for bert-base-multilingual-cased
|
12 |
+
|
13 |
+
Houlsby Adapter trained with Masked Language Modelling on Māori Wikipedia Articles for 50k steps and a batch size of 64.
|
14 |
+
|
15 |
+
|
16 |
+
**This adapter was created for usage with the [Adapters](https://github.com/Adapter-Hub/adapters) library.**
|
17 |
+
|
18 |
+
## Usage
|
19 |
+
|
20 |
+
First, install `adapters`:
|
21 |
+
|
22 |
+
```
|
23 |
+
pip install -U adapters
|
24 |
+
```
|
25 |
+
|
26 |
+
Now, the adapter can be loaded and activated like this:
|
27 |
+
|
28 |
+
```python
|
29 |
+
from adapters import AutoAdapterModel
|
30 |
+
|
31 |
+
model = AutoAdapterModel.from_pretrained("bert-base-multilingual-cased")
|
32 |
+
adapter_name = model.load_adapter("AdapterHub/bert-base-multilingual-cased-mi-wiki_houlsby")
|
33 |
+
model.set_active_adapters(adapter_name)
|
34 |
+
```
|
35 |
+
|
36 |
+
## Architecture & Training
|
37 |
+
|
38 |
+
- Adapter architecture: houlsby
|
39 |
+
- Prediction head: None
|
40 |
+
- Dataset: [mi/wiki](https://adapterhub.ml/explore/mi/wiki/)
|
41 |
+
|
42 |
+
## Author Information
|
43 |
+
|
44 |
+
- Author name(s): Jonas Pfeiffer
|
45 |
+
- Author email: [email protected]
|
46 |
+
- Author links: [Website](https://pfeiffer.ai), [GitHub](https://github.com/jopfeiff), [Twitter](https://twitter.com/@PfeiffJo)
|
47 |
+
|
48 |
+
## Versions
|
49 |
+
- `nd` **(main)**
|
50 |
+
- `wd`
|
51 |
+
|
52 |
+
## Citation
|
53 |
+
|
54 |
+
```bibtex
|
55 |
+
@article{pfeiffer20madx,
|
56 |
+
title={{MAD-X}: An {A}dapter-based {F}ramework for {M}ulti-task {C}ross-lingual {T}ransfer},
|
57 |
+
author={Pfeiffer, Jonas and Vuli\'{c}, Ivan and Gurevych, Iryna and Ruder, Sebastian},
|
58 |
+
journal={arXiv preprint},
|
59 |
+
year={2020},
|
60 |
+
url={https://arxiv.org/pdf/2005.00052.pdf},
|
61 |
+
}
|
62 |
+
|
63 |
+
```
|
64 |
+
|
65 |
+
*This adapter has been auto-imported from https://github.com/Adapter-Hub/Hub/blob/master/adapters/ukp/bert-base-multilingual-cased-mi-wiki_houlsby.yaml*.
|
adapter_config.json
ADDED
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"config": {
|
3 |
+
"adapter_residual_before_ln": false,
|
4 |
+
"cross_adapter": false,
|
5 |
+
"dropout": 0.0,
|
6 |
+
"factorized_phm_W": true,
|
7 |
+
"factorized_phm_rule": false,
|
8 |
+
"hypercomplex_nonlinearity": "glorot-uniform",
|
9 |
+
"init_weights": "bert",
|
10 |
+
"inv_adapter": "nice",
|
11 |
+
"inv_adapter_reduction_factor": 2,
|
12 |
+
"is_parallel": false,
|
13 |
+
"learn_phm": true,
|
14 |
+
"leave_out": [],
|
15 |
+
"ln_after": false,
|
16 |
+
"ln_before": false,
|
17 |
+
"mh_adapter": true,
|
18 |
+
"non_linearity": "gelu_new",
|
19 |
+
"original_ln_after": true,
|
20 |
+
"original_ln_before": false,
|
21 |
+
"output_adapter": true,
|
22 |
+
"phm_bias": true,
|
23 |
+
"phm_c_init": "normal",
|
24 |
+
"phm_dim": 4,
|
25 |
+
"phm_init_range": 0.0001,
|
26 |
+
"phm_layer": false,
|
27 |
+
"phm_rank": 1,
|
28 |
+
"reduction_factor": 2,
|
29 |
+
"residual_before_ln": true,
|
30 |
+
"scaling": 1.0,
|
31 |
+
"shared_W_phm": false,
|
32 |
+
"shared_phm_rule": true,
|
33 |
+
"use_gating": false
|
34 |
+
},
|
35 |
+
"hidden_size": 768,
|
36 |
+
"model_class": "BertAdapterModel",
|
37 |
+
"model_name": "bert-base-multilingual-cased",
|
38 |
+
"model_type": "bert",
|
39 |
+
"name": "mi",
|
40 |
+
"version": "0.2.0"
|
41 |
+
}
|
pytorch_adapter.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:23b866617a0b181c95ed6d1897de192f5704fa4bf04adb145571be23e7b4a0fc
|
3 |
+
size 57954050
|