update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-sa-4.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- te_dx_jp
|
| 7 |
+
model-index:
|
| 8 |
+
- name: t5-base-TEDxJP-1body-3context
|
| 9 |
+
results: []
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
+
|
| 15 |
+
# t5-base-TEDxJP-1body-3context
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [sonoisa/t5-base-japanese](https://huggingface.co/sonoisa/t5-base-japanese) on the te_dx_jp dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.4926
|
| 20 |
+
- Wer: 0.1968
|
| 21 |
+
- Mer: 0.1894
|
| 22 |
+
- Wil: 0.2793
|
| 23 |
+
- Wip: 0.7207
|
| 24 |
+
- Hits: 55899
|
| 25 |
+
- Substitutions: 6836
|
| 26 |
+
- Deletions: 3636
|
| 27 |
+
- Insertions: 2590
|
| 28 |
+
- Cer: 0.1733
|
| 29 |
+
|
| 30 |
+
## Model description
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Intended uses & limitations
|
| 35 |
+
|
| 36 |
+
More information needed
|
| 37 |
+
|
| 38 |
+
## Training and evaluation data
|
| 39 |
+
|
| 40 |
+
More information needed
|
| 41 |
+
|
| 42 |
+
## Training procedure
|
| 43 |
+
|
| 44 |
+
### Training hyperparameters
|
| 45 |
+
|
| 46 |
+
The following hyperparameters were used during training:
|
| 47 |
+
- learning_rate: 0.0001
|
| 48 |
+
- train_batch_size: 64
|
| 49 |
+
- eval_batch_size: 8
|
| 50 |
+
- seed: 42
|
| 51 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 52 |
+
- lr_scheduler_type: linear
|
| 53 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 54 |
+
- num_epochs: 10
|
| 55 |
+
|
| 56 |
+
### Training results
|
| 57 |
+
|
| 58 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer | Mer | Wil | Wip | Hits | Substitutions | Deletions | Insertions | Cer |
|
| 59 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:------:|:-----:|:-------------:|:---------:|:----------:|:------:|
|
| 60 |
+
| 0.7082 | 1.0 | 746 | 0.5637 | 0.2626 | 0.2430 | 0.3355 | 0.6645 | 54301 | 7195 | 4875 | 5358 | 0.2552 |
|
| 61 |
+
| 0.6213 | 2.0 | 1492 | 0.5150 | 0.2068 | 0.1994 | 0.2899 | 0.7101 | 55107 | 6861 | 4403 | 2462 | 0.1866 |
|
| 62 |
+
| 0.5331 | 3.0 | 2238 | 0.4945 | 0.2038 | 0.1958 | 0.2858 | 0.7142 | 55551 | 6845 | 3975 | 2705 | 0.1816 |
|
| 63 |
+
| 0.5185 | 4.0 | 2984 | 0.4880 | 0.2003 | 0.1929 | 0.2831 | 0.7169 | 55639 | 6860 | 3872 | 2563 | 0.1779 |
|
| 64 |
+
| 0.4963 | 5.0 | 3730 | 0.4858 | 0.1988 | 0.1912 | 0.2810 | 0.7190 | 55837 | 6838 | 3696 | 2662 | 0.1772 |
|
| 65 |
+
| 0.4625 | 6.0 | 4476 | 0.4885 | 0.1964 | 0.1894 | 0.2799 | 0.7201 | 55785 | 6875 | 3711 | 2448 | 0.1720 |
|
| 66 |
+
| 0.4416 | 7.0 | 5222 | 0.4898 | 0.1962 | 0.1890 | 0.2788 | 0.7212 | 55870 | 6819 | 3682 | 2522 | 0.1726 |
|
| 67 |
+
| 0.4287 | 8.0 | 5968 | 0.4894 | 0.1968 | 0.1894 | 0.2790 | 0.7210 | 55889 | 6804 | 3678 | 2580 | 0.1743 |
|
| 68 |
+
| 0.4457 | 9.0 | 6714 | 0.4909 | 0.1964 | 0.1891 | 0.2792 | 0.7208 | 55919 | 6858 | 3594 | 2586 | 0.1739 |
|
| 69 |
+
| 0.4068 | 10.0 | 7460 | 0.4926 | 0.1968 | 0.1894 | 0.2793 | 0.7207 | 55899 | 6836 | 3636 | 2590 | 0.1733 |
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
### Framework versions
|
| 73 |
+
|
| 74 |
+
- Transformers 4.12.5
|
| 75 |
+
- Pytorch 1.10.0+cu102
|
| 76 |
+
- Datasets 1.15.1
|
| 77 |
+
- Tokenizers 0.10.3
|