File size: 3,195 Bytes
bc44030
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: facebook/wav2vec2-base
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: nyankole_wav2vec2
  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. -->

# nyankole_wav2vec2

This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 2.8615
- Wer: 1.0

## 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: 0.0003
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 150
- num_epochs: 30
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step | Validation Loss | Wer |
|:-------------:|:-------:|:----:|:---------------:|:---:|
| 2.9097        | 0.9976  | 210  | 2.9031          | 1.0 |
| 2.8649        | 2.0     | 421  | 2.8534          | 1.0 |
| 2.8545        | 2.9976  | 631  | 2.8537          | 1.0 |
| 2.8465        | 4.0     | 842  | 2.8554          | 1.0 |
| 2.8509        | 4.9976  | 1052 | 2.8629          | 1.0 |
| 2.8491        | 6.0     | 1263 | 2.8828          | 1.0 |
| 2.8463        | 6.9976  | 1473 | 2.8570          | 1.0 |
| 2.8477        | 8.0     | 1684 | 2.8675          | 1.0 |
| 2.8478        | 8.9976  | 1894 | 2.8605          | 1.0 |
| 2.8411        | 10.0    | 2105 | 2.8593          | 1.0 |
| 2.8493        | 10.9976 | 2315 | 2.8573          | 1.0 |
| 2.8478        | 12.0    | 2526 | 2.8564          | 1.0 |
| 2.8823        | 12.9976 | 2736 | 2.8538          | 1.0 |
| 2.8413        | 14.0    | 2947 | 2.8534          | 1.0 |
| 2.8497        | 14.9976 | 3157 | 2.8487          | 1.0 |
| 2.8439        | 16.0    | 3368 | 2.8642          | 1.0 |
| 2.8442        | 16.9976 | 3578 | 2.8527          | 1.0 |
| 2.8425        | 18.0    | 3789 | 2.8611          | 1.0 |
| 2.841         | 18.9976 | 3999 | 2.8617          | 1.0 |
| 2.8426        | 20.0    | 4210 | 2.8563          | 1.0 |
| 2.8454        | 20.9976 | 4420 | 2.8527          | 1.0 |
| 2.8396        | 22.0    | 4631 | 2.8568          | 1.0 |
| 2.8449        | 22.9976 | 4841 | 2.8503          | 1.0 |
| 2.8424        | 24.0    | 5052 | 2.8596          | 1.0 |
| 2.8438        | 24.9976 | 5262 | 2.8624          | 1.0 |
| 2.8414        | 26.0    | 5473 | 2.8606          | 1.0 |
| 2.8387        | 26.9976 | 5683 | 2.8635          | 1.0 |
| 2.8408        | 28.0    | 5894 | 2.8569          | 1.0 |
| 2.8729        | 28.9976 | 6104 | 2.8640          | 1.0 |
| 2.8417        | 29.9287 | 6300 | 2.8615          | 1.0 |


### Framework versions

- Transformers 4.42.4
- Pytorch 2.1.0+cu118
- Datasets 2.20.0
- Tokenizers 0.19.1