File size: 1,517 Bytes
33d9861
7effc69
 
 
33d9861
774231b
 
 
 
 
 
33d9861
 
774231b
 
33d9861
774231b
33d9861
774231b
 
 
 
 
 
 
 
 
 
 
33d9861
774231b
33d9861
774231b
33d9861
774231b
33d9861
774231b
33d9861
774231b
33d9861
774231b
33d9861
774231b
33d9861
774231b
33d9861
774231b
 
 
 
 
 
 
 
 
 
 
 
33d9861
774231b
33d9861
774231b
 
 
 
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
---
base_model: facebook/mms-1b-all
datasets:
- common_voice_17_0
library_name: transformers
license: cc-by-nc-4.0
tags:
- generated_from_trainer
model-index:
- name: wav2vec2-mms-1b-CV17.0
  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. -->

# wav2vec2-mms-1b-CV17.0

This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the common_voice_17_0 dataset.
It achieves the following results on the evaluation set:
- eval_loss: 0.2564
- eval_wer: 0.4162
- eval_cer: 0.0698
- eval_bleu: 0.3568
- eval_runtime: 202.4945
- eval_samples_per_second: 11.921
- eval_steps_per_second: 1.491
- epoch: 4.2105
- step: 1200

## 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: 0.001
- 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_ratio: 0.15
- training_steps: 2000
- mixed_precision_training: Native AMP

### Framework versions

- Transformers 4.44.2
- Pytorch 2.4.0+cu121
- Datasets 3.0.0
- Tokenizers 0.19.1