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-8front-1body-8rear
|
| 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-8front-1body-8rear
|
| 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: 0.4359
|
| 20 |
+
- Wer: 0.1693
|
| 21 |
+
- Mer: 0.1635
|
| 22 |
+
- Wil: 0.2485
|
| 23 |
+
- Wip: 0.7515
|
| 24 |
+
- Hits: 55936
|
| 25 |
+
- Substitutions: 6242
|
| 26 |
+
- Deletions: 2409
|
| 27 |
+
- Insertions: 2283
|
| 28 |
+
- Cer: 0.1334
|
| 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: 32
|
| 49 |
+
- eval_batch_size: 32
|
| 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 |
+
| 0.5891 | 1.0 | 1457 | 0.4696 | 0.2065 | 0.1954 | 0.2843 | 0.7157 | 54931 | 6665 | 2991 | 3684 | 0.1713 |
|
| 61 |
+
| 0.4787 | 2.0 | 2914 | 0.4172 | 0.1767 | 0.1707 | 0.2566 | 0.7434 | 55432 | 6311 | 2844 | 2256 | 0.1417 |
|
| 62 |
+
| 0.4425 | 3.0 | 4371 | 0.4078 | 0.1727 | 0.1667 | 0.2528 | 0.7472 | 55754 | 6342 | 2491 | 2320 | 0.1362 |
|
| 63 |
+
| 0.3887 | 4.0 | 5828 | 0.4101 | 0.1718 | 0.1657 | 0.2521 | 0.7479 | 55850 | 6367 | 2370 | 2356 | 0.1346 |
|
| 64 |
+
| 0.3567 | 5.0 | 7285 | 0.4120 | 0.1708 | 0.1648 | 0.2505 | 0.7495 | 55926 | 6315 | 2346 | 2371 | 0.1333 |
|
| 65 |
+
| 0.3157 | 6.0 | 8742 | 0.4174 | 0.1710 | 0.1652 | 0.2510 | 0.7490 | 55805 | 6313 | 2469 | 2260 | 0.1346 |
|
| 66 |
+
| 0.2958 | 7.0 | 10199 | 0.4277 | 0.1700 | 0.1641 | 0.2493 | 0.7507 | 55934 | 6263 | 2390 | 2330 | 0.1339 |
|
| 67 |
+
| 0.3088 | 8.0 | 11656 | 0.4268 | 0.1687 | 0.1631 | 0.2483 | 0.7517 | 55890 | 6252 | 2445 | 2198 | 0.1329 |
|
| 68 |
+
| 0.2725 | 9.0 | 13113 | 0.4352 | 0.1697 | 0.1639 | 0.2490 | 0.7510 | 55911 | 6247 | 2429 | 2287 | 0.1329 |
|
| 69 |
+
| 0.2588 | 10.0 | 14570 | 0.4359 | 0.1693 | 0.1635 | 0.2485 | 0.7515 | 55936 | 6242 | 2409 | 2283 | 0.1334 |
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
### Framework versions
|
| 73 |
+
|
| 74 |
+
- Transformers 4.21.2
|
| 75 |
+
- Pytorch 1.12.1+cu116
|
| 76 |
+
- Datasets 2.4.0
|
| 77 |
+
- Tokenizers 0.12.1
|