File size: 3,949 Bytes
53d73b9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
license: apache-2.0
base_model: openai/whisper-small
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: Whisper Small superU
  results: []
---

<!-- 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 Small superU

This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 6.3374
- Wer: 57.2770

## 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: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 128
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 4000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer     |
|:-------------:|:------:|:----:|:---------------:|:-------:|
| 0.0           | 100.0  | 100  | 2.1912          | 59.6244 |
| 0.0           | 200.0  | 200  | 2.3111          | 58.2160 |
| 0.0           | 300.0  | 300  | 2.3886          | 58.2160 |
| 0.0           | 400.0  | 400  | 2.5319          | 60.5634 |
| 0.0           | 500.0  | 500  | 2.7160          | 60.0939 |
| 0.0           | 600.0  | 600  | 2.9609          | 61.0329 |
| 0.0           | 700.0  | 700  | 3.2141          | 61.0329 |
| 0.0           | 800.0  | 800  | 3.4591          | 61.5023 |
| 0.0           | 900.0  | 900  | 3.7213          | 62.4413 |
| 0.0           | 1000.0 | 1000 | 3.9804          | 61.0329 |
| 0.0           | 1100.0 | 1100 | 4.2234          | 78.4038 |
| 0.0           | 1200.0 | 1200 | 4.4138          | 63.3803 |
| 0.0           | 1300.0 | 1300 | 4.5889          | 77.9343 |
| 0.0           | 1400.0 | 1400 | 4.7946          | 70.8920 |
| 0.0           | 1500.0 | 1500 | 4.9337          | 65.7277 |
| 0.0           | 1600.0 | 1600 | 5.0758          | 56.8075 |
| 0.0           | 1700.0 | 1700 | 5.2692          | 56.8075 |
| 0.0           | 1800.0 | 1800 | 5.4087          | 56.8075 |
| 0.0           | 1900.0 | 1900 | 5.5500          | 56.8075 |
| 0.0           | 2000.0 | 2000 | 5.6783          | 56.8075 |
| 0.0           | 2100.0 | 2100 | 5.6287          | 56.8075 |
| 0.0           | 2200.0 | 2200 | 5.6852          | 56.3380 |
| 0.0           | 2300.0 | 2300 | 5.7374          | 56.3380 |
| 0.0           | 2400.0 | 2400 | 5.8023          | 56.3380 |
| 0.0           | 2500.0 | 2500 | 5.8672          | 57.2770 |
| 0.0           | 2600.0 | 2600 | 5.9427          | 57.2770 |
| 0.0           | 2700.0 | 2700 | 5.9891          | 57.2770 |
| 0.0           | 2800.0 | 2800 | 6.0490          | 57.2770 |
| 0.0           | 2900.0 | 2900 | 6.0639          | 57.2770 |
| 0.0           | 3000.0 | 3000 | 6.1095          | 57.2770 |
| 0.0           | 3100.0 | 3100 | 6.1477          | 57.2770 |
| 0.0           | 3200.0 | 3200 | 6.2039          | 57.2770 |
| 0.0           | 3300.0 | 3300 | 6.2346          | 57.2770 |
| 0.0           | 3400.0 | 3400 | 6.2567          | 57.2770 |
| 0.0           | 3500.0 | 3500 | 6.2841          | 57.2770 |
| 0.0           | 3600.0 | 3600 | 6.3028          | 57.2770 |
| 0.0           | 3700.0 | 3700 | 6.3029          | 57.2770 |
| 0.0           | 3800.0 | 3800 | 6.3294          | 57.2770 |
| 0.0           | 3900.0 | 3900 | 6.3346          | 57.2770 |
| 0.0           | 4000.0 | 4000 | 6.3374          | 57.2770 |


### Framework versions

- Transformers 4.44.2
- Pytorch 2.4.0+cu121
- Datasets 2.21.0
- Tokenizers 0.19.1