File size: 1,417 Bytes
017a22b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
---
license: cc
task_categories:
- sentence-similarity
language:
- ml
pretty_name: Malayalam ASR reference prediction pair
size_categories:
- n<1K
---
---
license: cc-by-4.0
task_categories:
- sentence-similarity
language:
- ml
pretty_name: Malayalam ASR Reference Prediction dataset
size_categories:
- n<1K
---
# Malayalam ASR Reference Prediction dataset

This repository contains evaluation results from the Malayalam ASR model "vrclc/Whisper_small_malayalam" using the "google/fleurs" dataset.

* ASR Model Name: [vrclc/Whisper_small_malayalam](https://huggingface.co/vrclc/Whisper_small_malayalam)
* Dataset: [google/fleurs](https://huggingface.co/datasets/google/fleurs/viewer/ml_in/test)


- **Curated by:** VRCLC
- [vrclc/Whisper_small_malayalam](https://huggingface.co/vrclc/Whisper_small_malayalam) was trained with 50 hours of Malayalam speech data.
- The test set of [google/fleurs](https://huggingface.co/datasets/google/fleurs/viewer/ml_in/test) dataset which consists of Malayalam speech data was used to evaluate the model
- The evaluation of 500 samples from [google/fleurs](https://huggingface.co/datasets/google/fleurs/viewer/ml_in/test) and gives a WER is 53% when evaluated after punctuation removal.




## Uses



### Direct Use

This repository serves to provide an insight in error analysis which helps to identify general mistakes and areas for improvement in Malayalam speech recognition.