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-1context
|
| 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-1context
|
| 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: 0.5061
|
| 20 |
+
- Wer: 0.1990
|
| 21 |
+
- Mer: 0.1913
|
| 22 |
+
- Wil: 0.2823
|
| 23 |
+
- Wip: 0.7177
|
| 24 |
+
- Hits: 55830
|
| 25 |
+
- Substitutions: 6943
|
| 26 |
+
- Deletions: 3598
|
| 27 |
+
- Insertions: 2664
|
| 28 |
+
- Cer: 0.1763
|
| 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 |
+
| 0.7277 | 1.0 | 746 | 0.5799 | 0.2384 | 0.2256 | 0.3188 | 0.6812 | 54323 | 7170 | 4878 | 3777 | 0.2371 |
|
| 61 |
+
| 0.6278 | 2.0 | 1492 | 0.5254 | 0.2070 | 0.1997 | 0.2905 | 0.7095 | 55045 | 6885 | 4441 | 2412 | 0.1962 |
|
| 62 |
+
| 0.5411 | 3.0 | 2238 | 0.5076 | 0.2022 | 0.1950 | 0.2858 | 0.7142 | 55413 | 6902 | 4056 | 2463 | 0.1805 |
|
| 63 |
+
| 0.53 | 4.0 | 2984 | 0.5020 | 0.1979 | 0.1911 | 0.2814 | 0.7186 | 55599 | 6849 | 3923 | 2362 | 0.1761 |
|
| 64 |
+
| 0.5094 | 5.0 | 3730 | 0.4999 | 0.1987 | 0.1915 | 0.2828 | 0.7172 | 55651 | 6944 | 3776 | 2465 | 0.1742 |
|
| 65 |
+
| 0.4783 | 6.0 | 4476 | 0.5016 | 0.1985 | 0.1914 | 0.2826 | 0.7174 | 55684 | 6947 | 3740 | 2490 | 0.1753 |
|
| 66 |
+
| 0.4479 | 7.0 | 5222 | 0.5035 | 0.1976 | 0.1905 | 0.2819 | 0.7181 | 55726 | 6961 | 3684 | 2468 | 0.1733 |
|
| 67 |
+
| 0.4539 | 8.0 | 5968 | 0.5022 | 0.1967 | 0.1896 | 0.2807 | 0.7193 | 55795 | 6938 | 3638 | 2477 | 0.1729 |
|
| 68 |
+
| 0.4632 | 9.0 | 6714 | 0.5034 | 0.1991 | 0.1913 | 0.2824 | 0.7176 | 55844 | 6942 | 3585 | 2687 | 0.1758 |
|
| 69 |
+
| 0.4201 | 10.0 | 7460 | 0.5061 | 0.1990 | 0.1913 | 0.2823 | 0.7177 | 55830 | 6943 | 3598 | 2664 | 0.1763 |
|
| 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
|