File size: 3,180 Bytes
b5a2c51
 
47efd17
 
b5a2c51
 
 
47efd17
 
 
b5a2c51
 
 
 
 
 
47efd17
b5a2c51
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
47efd17
b5a2c51
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
---
library_name: transformers
language:
- jv
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- whisper
- javanese
- asr
- generated_from_trainer
datasets:
- jv_id_asr_split
metrics:
- wer
model-index:
- name: Whisper-Tiny-Java-v3
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: jv_id_asr_split
      type: jv_id_asr_split
      config: jv_id_asr_source
      split: None
      args: jv_id_asr_source
    metrics:
    - name: Wer
      type: wer
      value: 0.2586507557925852
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# Whisper-Tiny-Java-v3

This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the jv_id_asr_split dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2980
- Wer: 0.2587

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 8
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- training_steps: 10000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step  | Validation Loss | Wer    |
|:-------------:|:------:|:-----:|:---------------:|:------:|
| 1.1788        | 0.0540 | 500   | 0.9671          | 0.6590 |
| 0.8015        | 0.1081 | 1000  | 0.6977          | 0.5305 |
| 0.6498        | 0.1621 | 1500  | 0.5725          | 0.6670 |
| 0.5828        | 0.2161 | 2000  | 0.5094          | 0.4829 |
| 0.5226        | 0.2702 | 2500  | 0.4642          | 0.3860 |
| 0.4955        | 0.3242 | 3000  | 0.4341          | 0.3915 |
| 0.4616        | 0.3782 | 3500  | 0.4128          | 0.3540 |
| 0.4474        | 0.4323 | 4000  | 0.3900          | 0.3614 |
| 0.4387        | 0.4863 | 4500  | 0.3736          | 0.3563 |
| 0.4154        | 0.5403 | 5000  | 0.3606          | 0.3274 |
| 0.419         | 0.5944 | 5500  | 0.3495          | 0.3144 |
| 0.3799        | 0.6484 | 6000  | 0.3398          | 0.2922 |
| 0.3802        | 0.7024 | 6500  | 0.3290          | 0.3044 |
| 0.3611        | 0.7565 | 7000  | 0.3225          | 0.2823 |
| 0.3548        | 0.8105 | 7500  | 0.3168          | 0.2733 |
| 0.346         | 0.8645 | 8000  | 0.3105          | 0.2660 |
| 0.3547        | 0.9186 | 8500  | 0.3063          | 0.2708 |
| 0.3211        | 0.9726 | 9000  | 0.3019          | 0.2827 |
| 0.2718        | 1.0267 | 9500  | 0.2990          | 0.2660 |
| 0.2859        | 1.0807 | 10000 | 0.2980          | 0.2587 |


### Framework versions

- Transformers 4.50.0.dev0
- Pytorch 2.6.0+cu126
- Datasets 3.4.0
- Tokenizers 0.21.1