File size: 1,572 Bytes
c47a3d8
 
dcd4d8f
 
 
 
 
 
 
 
 
c47a3d8
 
dcd4d8f
 
c47a3d8
dcd4d8f
c47a3d8
dcd4d8f
 
 
 
c47a3d8
dcd4d8f
c47a3d8
dcd4d8f
c47a3d8
dcd4d8f
c47a3d8
dcd4d8f
c47a3d8
dcd4d8f
c47a3d8
dcd4d8f
c47a3d8
dcd4d8f
c47a3d8
dcd4d8f
c47a3d8
dcd4d8f
 
 
 
 
 
 
 
 
 
 
 
c47a3d8
dcd4d8f
c47a3d8
dcd4d8f
 
 
 
c47a3d8
 
dcd4d8f
c47a3d8
dcd4d8f
 
 
 
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
---
library_name: transformers
license: mit
base_model: facebook/w2v-bert-2.0
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: W2V2-bert-Malasar
  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. -->

# W2V2-bert-Malasar

This model is a fine-tuned version of [facebook/w2v-bert-2.0](https://huggingface.co/facebook/w2v-bert-2.0) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 2.3938
- Wer: 1.0533

## 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: 5e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 10
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer    |
|:-------------:|:------:|:----:|:---------------:|:------:|
| No log        | 4.6512 | 100  | 3.1126          | 1.0    |
| No log        | 9.3023 | 200  | 2.3938          | 1.0533 |


### Framework versions

- Transformers 4.44.1
- Pytorch 2.0.1+cu117
- Datasets 2.16.0
- Tokenizers 0.19.1