File size: 1,725 Bytes
7a7c35f
 
 
 
0d927dc
7a7c35f
 
 
78ebb68
 
7a7c35f
 
 
78ebb68
7a7c35f
 
 
 
 
 
a8c37e9
7a7c35f
 
 
 
 
 
0d927dc
7a7c35f
 
 
 
 
 
 
0d927dc
 
7a7c35f
a8c37e9
7a7c35f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0d927dc
 
 
7a7c35f
 
 
 
 
 
 
0d927dc
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
---
language:
- ro
license: apache-2.0
base_model: iulik-pisik/horoscope_model_small
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- iulik-pisik/audio_vreme
metrics:
- wer
model-index:
- name: Horoscope Model Small - finetuned on weather
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: Vreme ProTV
      type: iulik-pisik/audio_vreme
      config: default
      split: test
      args: 'config: ro, split: test'
    metrics:
    - name: Wer
      type: wer
      value: 8.8779
---


# Horoscope Model Base - finetuned on weather

This model is a fine-tuned version of [iulik-pisik/horoscope_model_small](https://huggingface.co/iulik-pisik/horoscope_model_small) on the Vreme ProTV dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1875
- Wer: 8.8779

## Model description

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 3000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer     |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.0134        | 5.79  | 1000 | 0.1508          | 9.515   |
| 0.0006        | 11.82 | 2000 | 0.1713          | 9.1862  |
| 0.0001        | 17.39 | 3000 | 0.1875          | 8.8779  |


### Framework versions

- Transformers 4.39.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2