File size: 2,410 Bytes
6e56e8c 646fdc4 6e56e8c |
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 |
---
language:
- uk
license: apache-2.0
tags:
- automatic-speech-recognition
- mozilla-foundation/common_voice_7_0
- generated_from_trainer
- uk
- robust-speech-event
datasets:
- common_voice
model-index:
- name: wav2vec2-xls-r-300m-uk-with-lm
results: []
---
# Ukrainian STT model (with Language Model)
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_7_0 - UK dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3015
- Wer: 0.3377
- Cer: 0.0708
The above results present evaluation without the language model.
## Model description
On 100 test example the model shows the following results:
Without LM:
- WER: 0.2647058823529412
- CER: 0.046974185357596274
With LM:
- WER: 0.1568627450980392
- CER: 0.028988573846804908
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 20
- total_train_batch_size: 160
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 100.0
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
| 3.0255 | 7.93 | 500 | 2.5514 | 0.9921 | 0.9047 |
| 1.3809 | 15.86 | 1000 | 0.4065 | 0.5361 | 0.1201 |
| 1.2355 | 23.8 | 1500 | 0.3474 | 0.4618 | 0.1033 |
| 1.1956 | 31.74 | 2000 | 0.3617 | 0.4580 | 0.1005 |
| 1.1416 | 39.67 | 2500 | 0.3182 | 0.4074 | 0.0891 |
| 1.0996 | 47.61 | 3000 | 0.3166 | 0.3985 | 0.0875 |
| 1.0427 | 55.55 | 3500 | 0.3116 | 0.3835 | 0.0828 |
| 0.9961 | 63.49 | 4000 | 0.3137 | 0.3757 | 0.0807 |
| 0.9575 | 71.42 | 4500 | 0.2992 | 0.3632 | 0.0771 |
| 0.9154 | 79.36 | 5000 | 0.3015 | 0.3502 | 0.0740 |
| 0.8994 | 87.3 | 5500 | 0.3004 | 0.3425 | 0.0723 |
| 0.871 | 95.24 | 6000 | 0.3016 | 0.3394 | 0.0713 |
### Framework versions
- Transformers 4.16.0.dev0
- Pytorch 1.10.1+cu102
- Datasets 1.18.1.dev0
- Tokenizers 0.11.0
|