Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,33 @@
|
|
1 |
---
|
2 |
license: mit
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: mit
|
3 |
+
language:
|
4 |
+
- ko
|
5 |
+
metrics:
|
6 |
+
- cer
|
7 |
+
pipeline_tag: automatic-speech-recognition
|
8 |
+
tags:
|
9 |
+
- ksponspeech
|
10 |
+
model-index:
|
11 |
+
- name: cwwojin/stt_kr_conformer_ctc_medium
|
12 |
+
results:
|
13 |
+
- task:
|
14 |
+
type: automatic-speech-recognition # Required. Example: automatic-speech-recognition
|
15 |
+
dataset:
|
16 |
+
type: Murple/ksponspeech # Required. Example: common_voice. Use dataset id from https://hf.co/datasets
|
17 |
+
name: KsponSpeech (Korean) # Required. A pretty name for the dataset. Example: Common Voice (French)
|
18 |
+
split: test # Optional. Example: test
|
19 |
+
metrics:
|
20 |
+
- type: cer # Required. Example: wer. Use metric id from https://hf.co/metrics
|
21 |
+
value: 0.0 # Required. Example: 20.90
|
22 |
+
name: Test CER (%) # Optional. Example: Test WER
|
23 |
---
|
24 |
+
|
25 |
+
# stt_kr_conformer_ctc_small_10
|
26 |
+
- Fine-tuned from "stt_en_conformer_ctc_small"
|
27 |
+
- Trained on KsponSpeech, provided by https://aihub.or.kr/
|
28 |
+
## Preprocessing
|
29 |
+
- Files converted from .pcm -> .wav
|
30 |
+
- Text - Korean phonetic transcription
|
31 |
+
- SentencePiece tokenizer (Byte-pair encoding), vocab-size = 5,000
|
32 |
+
## Evaluation
|
33 |
+
- "KsponSpeech_eval_clean", "KsponSpeech_eval_other
|