Downloads · 30 days
109
5% of all-time downloads
wav2vec2-xls-r-300m-japanese
vitouphy/wav2vec2-xls-r-300m-japanese
wav2vec2-xls-r-300m-japanese is a automatic speech recognition model from vitouphy. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.
All-time downloads
2.2K
Public
Repo size
50.4 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.binary2.1 GB · 62%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- License
- apache-2.0
- Model type
- wav2vec2
- Access
- Public
- Created
- Mar 2, 2022
- Updated
- Mar 23, 2022
- SHA
- 6ca1b5ac
Model card
From the Hugging Face model README
This model is for transcribing audio into Hiragana, one format of Japanese language.
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the mozilla-foundation/common_voice_8_0 dataset. Note that the following results are achieved by:
- Modify
eval.pyto suit the use case. - Since kanji and katakana shares the same sound as hiragana, we convert all texts to hiragana using pykakasi and tokenize them using fugashi.
It achieves the following results on the evaluation set:
- Loss: 0.7751
- Cer: 0.2227
Evaluation results (Running ./eval.py):
| Model | Metric | Common-Voice-8/test | speech-recognition-community-v2/dev-data |
|---|---|---|---|
| w/o LM | WER | 0.5964 | 0.5532 |
| CER | 0.2944 | 0.2629 | |
| w/ LM | WER | 0.5405 | 0.4877 |
| CER | 0.2754 | 0.2487 |
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- 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: 1000
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 4.4081 | 1.6 | 500 | 4.0983 | 1.0 |
| 3.303 | 3.19 | 1000 | 3.3563 | 1.0 |
| 3.1538 | 4.79 | 1500 | 3.2066 | 0.9239 |
| 2.1526 | 6.39 | 2000 | 1.1597 | 0.3355 |
| 1.8726 | 7.98 | 2500 | 0.9023 | 0.2505 |
| 1.7817 | 9.58 | 3000 | 0.8219 | 0.2334 |
| 1.7488 | 11.18 | 3500 | 0.7915 | 0.2222 |
| 1.7039 | 12.78 | 4000 | 0.7751 | 0.2227 |
| Stop & Train | ||||
| 1.6571 | 15.97 | 5000 | 0.6788 | 0.1685 |
| 1.520400 | 19.16 | 6000 | 0.6095 | 0.1409 |
| 1.448200 | 22.35 | 7000 | 0.5843 | 0.1430 |
| 1.385400 | 25.54 | 8000 | 0.5699 | 0.1263 |
| 1.354200 | 28.73 | 9000 | 0.5686 | 0.1219 |
| 1.331500 | 31.92 | 10000 | 0.5502 | 0.1144 |
| 1.290800 | 35.11 | 11000 | 0.5371 | 0.1140 |
| Stop & Train | ||||
| 1.235200 | 38.30 | 12000 | 0.5394 | 0.1106 |
Framework versions
- Transformers 4.17.0.dev0
- Pytorch 1.10.2+cu102
- Datasets 1.18.2.dev0
- Tokenizers 0.11.0
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- wav2vec2
- License
- apache-2.0
- Languages
- ja
- Created
- Mar 2, 2022
- Updated
- Mar 23, 2022