patrickvonplaten
commited on
Commit
•
9a49133
1
Parent(s):
7c13d6f
Update README.md
Browse files
README.md
CHANGED
@@ -12,4 +12,22 @@ pipeline_tag: automatic-speech-recognition
|
|
12 |
license: apache-2.0
|
13 |
---
|
14 |
|
15 |
-
# Wav2Vec2-XLS-R-300M-21-EN
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
12 |
license: apache-2.0
|
13 |
---
|
14 |
|
15 |
+
# Wav2Vec2-XLS-R-300M-21-EN
|
16 |
+
|
17 |
+
Facebook's Wav2Vec2 XLS-R fine-tuned for Speech Translation.
|
18 |
+
|
19 |
+
This is a [SpeechEncoderDecoderModel](https://huggingface.co/transformers/model_doc/speechencoderdecoder.html) model.
|
20 |
+
The encoder was warm-started from the [`facebook/wav2vec2-xls-r-300m`](https://huggingface.co/facebook/wav2vec2-xls-r-300m) checkpoint and
|
21 |
+
the decoder was warm-started from the [`facebook/mbart-large-50`](https://huggingface.co/facebook/mbart-large-50) checkpoint.
|
22 |
+
Consequently, the encoder-decoder model was fine-tuned on 21 {lang}-to-English translation pairs of the [Covost2 dataset](https://huggingface.co/datasets/covost2).
|
23 |
+
|
24 |
+
For more information, please refer to Section *5.1.2* of the [official XLS-R paper]( ).
|
25 |
+
|
26 |
+
## Usage
|
27 |
+
|
28 |
+
TODO...
|
29 |
+
|
30 |
+
## Results
|
31 |
+
|
32 |
+
TODO...
|
33 |
+
|