Add pipeline tag, library name and link to code (#1)
Browse files- Add pipeline tag, library name and link to code (924d2fff372eb09491a2a618e871c674bdf0b10f)
Co-authored-by: Niels Rogge <nielsr@users.noreply.huggingface.co>
README.md
CHANGED
|
@@ -1,19 +1,21 @@
|
|
| 1 |
---
|
| 2 |
-
license: cc-by-4.0
|
| 3 |
-
language:
|
| 4 |
-
- en
|
| 5 |
-
- it
|
| 6 |
datasets:
|
| 7 |
- FBK-MT/mosel
|
| 8 |
- facebook/covost2
|
| 9 |
- openslr/librispeech_asr
|
| 10 |
- facebook/voxpopuli
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
metrics:
|
| 12 |
- wer
|
| 13 |
tags:
|
| 14 |
- speech
|
| 15 |
- speech recognition
|
| 16 |
- ASR
|
|
|
|
|
|
|
| 17 |
---
|
| 18 |
|
| 19 |
# FAMA-medium-asr
|
|
@@ -40,7 +42,6 @@ All the artifacts used for realizing FAMA models, including codebase, datasets,
|
|
| 40 |
themself are [released under OS-compliant licenses](#license), promoting a more
|
| 41 |
responsible creation of models in our community.
|
| 42 |
|
| 43 |
-
|
| 44 |
It is available in 2 sizes, with 2 variants for ASR only:
|
| 45 |
|
| 46 |
- [FAMA-small](https://huggingface.co/FBK-MT/fama-small) - 475 million parameters
|
|
@@ -50,6 +51,7 @@ It is available in 2 sizes, with 2 variants for ASR only:
|
|
| 50 |
|
| 51 |
For more information about FAMA, please check our [blog post](https://huggingface.co/blog/FAMA/release) and the [arXiv](https://arxiv.org/abs/2505.22759) preprint.
|
| 52 |
|
|
|
|
| 53 |
|
| 54 |
## Usage
|
| 55 |
|
|
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
datasets:
|
| 3 |
- FBK-MT/mosel
|
| 4 |
- facebook/covost2
|
| 5 |
- openslr/librispeech_asr
|
| 6 |
- facebook/voxpopuli
|
| 7 |
+
language:
|
| 8 |
+
- en
|
| 9 |
+
- it
|
| 10 |
+
license: cc-by-4.0
|
| 11 |
metrics:
|
| 12 |
- wer
|
| 13 |
tags:
|
| 14 |
- speech
|
| 15 |
- speech recognition
|
| 16 |
- ASR
|
| 17 |
+
library_name: fairseq
|
| 18 |
+
pipeline_tag: automatic-speech-recognition
|
| 19 |
---
|
| 20 |
|
| 21 |
# FAMA-medium-asr
|
|
|
|
| 42 |
themself are [released under OS-compliant licenses](#license), promoting a more
|
| 43 |
responsible creation of models in our community.
|
| 44 |
|
|
|
|
| 45 |
It is available in 2 sizes, with 2 variants for ASR only:
|
| 46 |
|
| 47 |
- [FAMA-small](https://huggingface.co/FBK-MT/fama-small) - 475 million parameters
|
|
|
|
| 51 |
|
| 52 |
For more information about FAMA, please check our [blog post](https://huggingface.co/blog/FAMA/release) and the [arXiv](https://arxiv.org/abs/2505.22759) preprint.
|
| 53 |
|
| 54 |
+
Code is available at https://github.com/hlt-mt/FBK-fairseq
|
| 55 |
|
| 56 |
## Usage
|
| 57 |
|