File size: 2,040 Bytes
749a73a
 
 
 
 
 
7ad781c
 
749a73a
 
 
 
7ad781c
 
 
 
 
 
 
 
 
 
 
749a73a
 
 
 
 
 
 
7ad781c
749a73a
7ad781c
 
749a73a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
library_name: transformers
license: apache-2.0
base_model: openai/whisper-medium
tags:
- generated_from_trainer
datasets:
- lozgen
metrics:
- wer
model-index:
- name: whisper-medium-lozgen-male-model
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: lozgen
      type: lozgen
    metrics:
    - name: Wer
      type: wer
      value: 0.4796960341961529
---

<!-- 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-medium-lozgen-male-model

This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the lozgen dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7984
- Wer: 0.4797

## 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: 1e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Use 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_steps: 500
- training_steps: 5000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step | Validation Loss | Wer    |
|:-------------:|:-------:|:----:|:---------------:|:------:|
| 1.2961        | 2.9851  | 200  | 0.7984          | 0.4797 |
| 0.2221        | 5.9701  | 400  | 0.8183          | 0.4540 |
| 0.0963        | 8.9552  | 600  | 0.8391          | 0.3890 |
| 0.0457        | 11.9403 | 800  | 0.8436          | 0.3887 |


### Framework versions

- Transformers 4.47.1
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.0