Add pipeline tag and library name
#1
by nielsr HF Staff - opened
README.md
CHANGED
|
@@ -1,6 +1,14 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
| 2 |
language:
|
| 3 |
- sk
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
tags:
|
| 5 |
- speech
|
| 6 |
- asr
|
|
@@ -9,11 +17,6 @@ tags:
|
|
| 9 |
- parliament
|
| 10 |
- legal
|
| 11 |
- politics
|
| 12 |
-
base_model: openai/whisper-small
|
| 13 |
-
datasets:
|
| 14 |
-
- erikbozik/slovak-plenary-asr-corpus
|
| 15 |
-
metrics:
|
| 16 |
-
- wer
|
| 17 |
model-index:
|
| 18 |
- name: whisper-small-sk
|
| 19 |
results:
|
|
@@ -24,9 +27,9 @@ model-index:
|
|
| 24 |
name: Common Voice 21 (Slovak test set)
|
| 25 |
type: common_voice
|
| 26 |
metrics:
|
| 27 |
-
-
|
| 28 |
-
type: wer
|
| 29 |
value: 25.7
|
|
|
|
| 30 |
- task:
|
| 31 |
type: automatic-speech-recognition
|
| 32 |
name: Automatic Speech Recognition
|
|
@@ -34,15 +37,15 @@ model-index:
|
|
| 34 |
name: FLEURS (Slovak test set)
|
| 35 |
type: fleurs
|
| 36 |
metrics:
|
| 37 |
-
-
|
| 38 |
-
type: wer
|
| 39 |
value: 10.6
|
| 40 |
-
|
| 41 |
---
|
| 42 |
|
| 43 |
# Whisper Small — Fine-tuned on Slovak Plenary ASR Corpus
|
| 44 |
|
| 45 |
-
This model is a fine-tuned version of [`openai/whisper-small`](https://huggingface.co/openai/whisper-small).
|
|
|
|
| 46 |
It is adapted for **Slovak ASR** using [SloPalSpeech](https://huggingface.co/datasets/erikbozik/slovak-plenary-asr-corpus): **2,806 hours** of aligned, ≤30 s speech–text pairs from official plenary sessions of the **Slovak National Council**.
|
| 47 |
|
| 48 |
- **Language:** Slovak
|
|
|
|
| 1 |
---
|
| 2 |
+
base_model: openai/whisper-small
|
| 3 |
+
datasets:
|
| 4 |
+
- erikbozik/slovak-plenary-asr-corpus
|
| 5 |
language:
|
| 6 |
- sk
|
| 7 |
+
license: mit
|
| 8 |
+
metrics:
|
| 9 |
+
- wer
|
| 10 |
+
library_name: transformers
|
| 11 |
+
pipeline_tag: automatic-speech-recognition
|
| 12 |
tags:
|
| 13 |
- speech
|
| 14 |
- asr
|
|
|
|
| 17 |
- parliament
|
| 18 |
- legal
|
| 19 |
- politics
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
model-index:
|
| 21 |
- name: whisper-small-sk
|
| 22 |
results:
|
|
|
|
| 27 |
name: Common Voice 21 (Slovak test set)
|
| 28 |
type: common_voice
|
| 29 |
metrics:
|
| 30 |
+
- type: wer
|
|
|
|
| 31 |
value: 25.7
|
| 32 |
+
name: WER
|
| 33 |
- task:
|
| 34 |
type: automatic-speech-recognition
|
| 35 |
name: Automatic Speech Recognition
|
|
|
|
| 37 |
name: FLEURS (Slovak test set)
|
| 38 |
type: fleurs
|
| 39 |
metrics:
|
| 40 |
+
- type: wer
|
|
|
|
| 41 |
value: 10.6
|
| 42 |
+
name: WER
|
| 43 |
---
|
| 44 |
|
| 45 |
# Whisper Small — Fine-tuned on Slovak Plenary ASR Corpus
|
| 46 |
|
| 47 |
+
This model is a fine-tuned version of [`openai/whisper-small`](https://huggingface.co/openai/whisper-small) presented in the paper [SloPal: A 60-Million-Word Slovak Parliamentary Corpus with Aligned Speech and Fine-Tuned ASR Models](https://huggingface.co/papers/2509.19270).
|
| 48 |
+
|
| 49 |
It is adapted for **Slovak ASR** using [SloPalSpeech](https://huggingface.co/datasets/erikbozik/slovak-plenary-asr-corpus): **2,806 hours** of aligned, ≤30 s speech–text pairs from official plenary sessions of the **Slovak National Council**.
|
| 50 |
|
| 51 |
- **Language:** Slovak
|